Effect of oral contraceptive metabolism of some trace Sheldon
Mwgen,
M.D.
and Janet
C. King,
agents
on the 2
Ph.D.
alterations noted in women taking the oral in blood levels of some trace minerals. Only a few of these, namely, iron, copper and zinc, have been studied. This paper intends to review the changes reported and outline the beginning of a series of studies to attempt to determine whether there is any metabolic significance to these alterations. The alterations reported in blood levels are generally believed to be related in large measure to alterations in levels of specific “transport” proteins induced primarily by estrogen. The significance of these changes is generally unknown, however, they involve slight elevations in serum iron and copper and possibly a reduction in serum zinc, although these are not as clear-cut as those of the other two trace minerals. In an attempt to determine whether any metabolic significance could be ABSTRACT contraceptive
Biochemical and agents have included
attributed
to
these
changes,
metabolic changes
several
studies
have
been
instituted.
The
first,
reported
in
this
paper, consisted of a metabolic study on a group of seven women aged 18-32 who had been taking oral contraceptives for at least 2 years. The subjects were studied over four metabolic periods for a total length of 1 16 days. They were studied in a random fashion, generally 2 months on and 2 months off the pill, although this varied slightly. The subjects were fed a controlled formula diet throughout this study and vitamins and trace minerals, other than copper, iron and zinc, were given orally in capsules. A trace mineral solution was administered so that the total daily iron, copper and zinc could be fed in four equal feedings per day with the formula. Measurements consisted of studies of the blood levels of the trace minerals as well as their binding proteins, various indices of blood chemistry to assess health, hematological indices and total balance studies for each of the trace minerals and nitrogen. The balance studies included not only urine and fecal loss but also skin, menstrual and hair losses. Preliminary results are discussed as well as the formidable problems involved in trace mineral studies in humans. These problems are not only related to contamination but also are analytic in nature. Am. J. Clin. Nutr. 28: 392-402, 1975.
In general, extreme paucity contraceptive lism. Most
of the zinc.
minerals, papers
contraceptive levels of alterations
can be said that there is an of data regarding the effects of
steroids on trace of the reports deal
trace The
it
agents these are
namely, report can
mineral metabowith only a few
iron,
that cause
trace mineral generally due
copper,
certain changes
the
in blood These in the
of
laneous
In
about
capacity
levels of the specific “transport” proteins. The significance, if any, of these changes is generally unknown.
(2,
More
recently,
attempts
have
been
392
The American
Journal
of Clinical
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
above
trace
substances
minerals
and
follows
1947
the
first
demonstrated that pregnancy increase on levels of iron and somewhat contraceptive elevating
made to ascertain which of the hormonal agents in the contraceptive steroids are responsible for the observed changes. There is virtually no work that I am aware of that has examined other aspects of trace mineral metabolism in relation to the contraceptive steroids. A brief review of the significant findings relating to
the related
miscel-
below.
Iron
and of
substances. to changes
each
that
(1).
These later
conclusive
observations
were
the observation pills likewise have an serum iron and iron binding
was
serum
iron
Sciences, 94720. 2This research was supported in part from the National Institutes of Health, 72-2790. 28: APRIL
1975,
that the effect on capacity
observation
the
1 Department of Nutritional California, Berkeley, California
Nutrition
followed
by
3). The important even though
pp. 392-402.
study
had a striking total iron binding
made
binding
University by a contract NIH-NICHD
Printed
in U.S.A.
of
EFFECT
capacity
OF
ORAL
CONTRACEPTIVE
is that
might be strikingly elevated in women the pill, there was no significant in hemoglobin or mean corpuscular
receiving
change hemoglobin
concentration
possible
increase
However, studies globin concentration
was,
in women receiving -the intrauterine measure
(2).
and
only
in packed cell did demonstrate on
a
whole,
slight
higher
the pill than those utilizing device as a contraceptive
In
a
comparison
between
the
effects of pregnancy and combined or sequential therapy on iron or iron binding capacity, it was found that the major increases in pregnancy occur during the second and third trimesters. Almost all of the increase was associated with an increase in transferrin (3). It was pointed out that these changes in total iron binding
capacity
(TIBC)
or
transferrin
ON
could
METABOLISM
of Norby
(7),
of a combined absorption by
volume (1). that hemo-
the
AGENTS
who
oral use of
393
studied
the influence
contraceptive radioiron.
on iron women
Six
were studied. Two absorption studies were made before oral contraceptives were administered. The first was a control and the second after a phlebotomy of 400 ml of blood. The second series of studies was similar to the first and was repeated after 3 months on oral contraceptive pills. (The removed blood was reinfused into the patients after absorption studies
were
basal
absorption
were
the
complete.)
before
same,
and
after
increased absorption was noted on and Plasma
iron
and
The
results
showed
and
during
therapy
the
of off
phlebotomy,
an
the same magnitude oral contraceptives.
TIBC
increased
during
oral
occur very rapidly, within 24 hours of institution of treatment with the contraceptive drugs.
contraceptive therapy. However, it is interesting to note that this increased serum iron and TIBC did not prevent the fall in hemoglobin after the
This
phlebotomy
an
increase increase
was
contraceptive delayed rapidly.
followed iron.
agent,
been
menstrual
the
recognized cycle
in
10 days
of the
TIBC
and
then
preliminary
some
associated
fell
contraceptives mia exhibit
that
with
the
fluctua-
tions in iron and iron binding capacities. Both the serum iron and iron binding capacity were lowest at, or immediately after, the onset of the menses when expressed as deviations from normal. Although it was observed that the mean level of iron and iron binding capacity were higher in individuals on the pill, the fluctuation within the cycle was not abolished by the administration of oral contraceptives (5).
It has
been
pointed
out
that
the
was
in
observa-
time
and
response
was
effect was from the the estrogens (4). It
for was
by
cessation
fall
from
principal rather than
immediately
After
the
7 to
for about It appeared
tions that progestogens had
not
in serum
absorption
part
no
noted.
the
change This
in
reticulocyte
finding
observation
that
who develop only a minimal
helps
women
explain
on
iron-deficiency decrease
oral
anein serum
iron. More recently pure progestogen
it has been materials
demonstrated failed to affect
(8).
that the
compound a substance is not metabolized to estrogen, was to be ineffective in changing the serum
serum levels of transferrin employed , medroxyprogesterone
which shown
The ,
levels. The
possibility
remains
separate mechanisms iron binding capacities. observed attributed
on both entirely
transferrin
that
affecting Most
there
may
iron and TIBC have to effect of the hormones
elevation.
Zilva
be
iron levels and of the changes
(9)
been on
suggests
a dis-
of iron seems to be related to the percent saturation of transferrin. The administration of oral contraceptives increased serum transferrin and TIBC to levels higher than those observed in iron-deficiency anemia. Even though iron levels are increased above normal, the percent saturation is less than that observed in the normal individual. The possibility therefore
iron binding capacity; 2) in early pregnancy there is a rise in iron levels prior to the elevation of iron binding capacity; and 3) in women on contraceptive pills the total iron binding capacity may rise without significant
exists
elevation
that
this
alteration
capacity-serum
iron
in total
relationship
iron
binding
has
attempted
to
answer
these
questions
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
must
changes
cycle
(% saturation)
may have an effect on iron transport by the gut (6). The only research I am acquainted with which
sociation
in
iron
cannot
levels
because
during
be accounted
of
these
British Medical the contraceptive
device supplying
considered
the
1)
the
menstrual
for by alterations
in
of iron.
In view the that
be
for
women.
sufficient
considerations, Journal pill
In view iron
the
editor
of
(10) has suggested may be a protective
of the to
women
problem in
of the
394
MARGEN
reproductive years, the contraceptive effect
in
conserving
nisms whereby the decrease the pill percent
iron
iron the
second
With
the
trend
with
or
one iron
of
to date
been estrogenic
of
the
potency.
In
Vrettos
et
an effect on the In pregnancy, excretion but are excretion
of
unsaturated
pregnancy
was
iron
noted.
invalidate the are primarily responsible but certainly shows that
metabolite
products
does
not
iron the
obvi-
have
a
diurnal
fluctuations
commented
that
the
early in afternoon.
on by Ramsey
plasma
iron
were
(1 2) who
reached
its
the morning and This phenomenon
the
These
stated
highest
its lowest was more
level in the recently
reinvestigated by Wiltnik et al. (13) on people eating a normal diet. They show that in the normal
free-living
variation occurring
and
in in
the
morning. reached is set
at
65-75%
plasma iron, the afternoon
lowest In
individual,
this
at
there the highest and early
night
study
is
and
the
early
lowest
values
great
values evening
in
the were
at about 4:00 AM. If the lowest value 0%, the peak values are as much as higher, reaching and gradually
4:00
PM
What
is most
interesting
these decreasing
about
levels around thereafter.
this
observation
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
be
influenced
very
markedly
were
decreased
of
Transferrin
.
transferrin
used on
crude
been well resulted
especially mestens
was total was
time to be essentially unchanged. pointed out that the method electrophoresis
It had pregnancy
with
first
in combination
as present only in pregnancy appear. Serum albumin and
gel was a relatively
capacity.
iron.
or
in Enovid) had marked of the serum proteins.
were
determination
binding
in plasma
that
workers,
in iron must be and iron
do
Proteins
should
that for this one of the
to
alone
methylester on various
out
estradiol,
haptoglobins said at that
no
correlation with the changes capacity. A word of caution into any studies of iron has
administered
(as its influence
pointed
ethinyl
varies
significant binding introduced
This
whether
et al. (14)
estrogen
by Smithies noted to
binding
to
Musa
values of the
notion
appear
as 1967 synthetic
serum it is
This
not
early the
were G0-globulin, ceruloplasmin, thyroid binding globulins and corticosteroid binding globulins. A so-called “pregnancy protein” described
whether
from 4.7 to 25 mg/24 hours. However, correlation was noted between the total binding capacity and estriol levels, although and
that
agents.
determine
estriol
As
continue effects
entirely clear are due to
urinary estrogens
The
the
effects. to
lowest
does
it is
Copper
activity
newer
still not observed
estrogenic
out, the estriol highest of all
ously estrogens effect,
Hence,
attempted.
agents, questions magnitude of
pointed are the
capacity
period.
tionships.
has
attempted
in total,
24-hour
no
However,
theory
the estrogens has binding capacity.
elevation
the
in on
it is effects
(1 1) have
over
possible effects changes of TIBC may have
some
progestational al.
transferrin
quite clear that when measuring serum iron, one must take the measurements at the same time of day and consider the dietary interrela-
to decrease
instances, the
mechaare with
of oral contraceptive to arise as to the many whether
The
brought about loss associated
absorption.
of
observed
be
KING
is that these changes in iron levels are not associated with corresponding levels in transferrin. There is less than a 10% variation in
exists that a beneficial
stores.
this may in menstrual
and the saturation
increasing test
the possibility pill may have
AND
during (15) and
It for
by
these
polyacrylyamide
technique.
known for in increase
many years that in copper levels,
the second as early as
and 1956
third
tn-
Russ and Raymunt (1 6) had demonstrated that estrogen would produce an increase in serum copper and copper
al.
binding
(17)
copper
oral
protein.
and
the
1966
in copper
sequential
or
There did excretion;
mixed
in
receiving
was
a
levels
a twofold increase 1 7- 1 8 days of the
et
increase
in subjects
There
increase
Carruthers
marked
ceruloplasmin
contraceptives.
threefold mately within
In
demonstrated
and
two-
to
approxi-
in cenuloplasmin onset of either
contraceptive
therapy.
not appear to be any effect both controls and individuals
on
urine receiv-
ing the material were said to be excreting less than 0.1 mg/liter of copper. The rise in serum concentration has been confirmed by other workers(18, 19). More
recently
Bniggs
et
al.
that progestational agents on serum copper, whereas has a marked effect on increasing shown effect
In fact,
the
increase
in serum
(18,
20)
have
alone have no estrogen alone serum copper.
copper
appears
to
EFFECT
OF ORAL
CONTRACEPTIVE
be proportional to the dose of estrogen administered. It is rather interesting to note that the serum copper and ceruloplasmin levels in pregnancy tend to reach their maximum at about the 22nd to 25th week of gestation and then gradually fall to the time of delivery (21). More recently, Daunter and Elstein (22) have evaluated the levels of serum copper, ceruloplasmin
and
ceruloplasmin
oxidase
activity
in
women using copper-containing intrauterine devices and in women taking combined oral contraceptives. No effect on serum copper or ceruloplasmin levels was noted in women wearing the intrauterine device. The levels reported although
were well no normal
within controls
study. The women agents were of two 1 .0 mg Norethisterone
receiving groups. plus
tradiol
and
Norethisterone ethinylestradiol. ceruloplasmin tivity were suggests
that
the
other,
half
the normal were used
oral contraceptive One group received SO zg of ethinylesthe
with the same No differences and ceruloplasmin noted in these two the
effect
was
limits, in this
almost
amount
of
amount of in copper, oxidase acgroups. This entirely
due
to the estrogen. Schenker et al. (23) have also studied a group of women taking the oral contraceptive agents with and without choleostasis. The oral contraceptive agents demonstrate
the
usual
increase
in copper
and
cerulo-
plasmin. Also, correlation coefficients between copper and ceruloplasmin is extremely high, i.e., increases in the patients receiving oral contraceptives (r = 0.88) versus the controls (r = 0.74). Patients who develop choleostasis while on the oral contraceptive agents have even higher copper and ceruloplasmin, with very high correlation (r 0.94) when this condition is superimposed. Olatunbosun
et
al.
(24)
have
studied
the
effect of oral contraceptives on serum copper levels in Nigerian women. In this study they were interested in the suggestion that thromboembolic phenomena are rarer in black United States females than in white (25). They studied patients with the intrauterine device in place versus women on two different oral contraceptive agents. As noted in prior studies, the intrauterine device had no effect on serum copper levels although the oral contraceptive agents produced the expected rise in copper. However, in general, the degree of elevation was
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
AGENTS lower
in
ON METABOLISM this
been 30%. The reason for this is unclear, although as pointed out, women in Nigeria frequently have lower serum albumin and also frequently reported,
group
i.e.,
elevated
an
than
395
increase
7-globulin.
had
of
Whether
usually
only
about
these
lesser
eleva-
tions are due to changes in serum protein levels due to genetic or nutritional differences, is also quite unclear. From a metabolic point of view, the significance of these changes in copper and ceruloplasmin levels are totally unknown and quite confusing. This is particularly true if one accepts the suggestion that ceruloplasmin is not considered to be a copper transport protein in the manner of transferrin (26). In fact, it is suggested that the metabolically active and exchangeable copper, i.e., that which is derived from dietary sources and other free copper, is principally transported by albumin and amino acids.
The
latter
copper
is in equilibrium
with
a
pool of “free” copper in the erythrocytes (27). More recently, a number of copper-binding proteins have been described but their activity, as well as the general action of the free or poorly bound copper, is largely a mystery at this time. It appears that one of the major functions of ceruloplasmin may be related to the release of bound iron but what regulates changes in the levels of this circulating enzyme, is unknown. However, the possibility exists that in pregnancy, when there is a great demand for transfer
of
iron,
copper
and
across the placental membrane, these transport systems may induced.
other
materials
mobilization be hormonally
of
Zinc
Several observations have been made regarding zinc levels and to date there does not appear to be a universal agreement as to the effects of the steroids on this mineral. Halstead in 1968 (28)
stated
that
both
pregnancy
and
oral
contraceptive agents levels although he
cause a fall in serum zinc confirmed the elevation
previously
in
observed
copper
levels
in
these
two conditions. In the following year, O’Leary and Spellacy (29) found that although there seemed to be a slightly lowered zinc level in pregnancy this was not statistically significant, although the oral contraceptive agents which he employed markedly elevated the zinc levels. More
recently
these
observations
of
O’Leary
396
MARGEN
AND
have been confirmed by reported measurements of Schenker et al. (30). The above observations point out the care which must be used
in
interpreting
of
the
reported
these materials. Again, care must zinc levels because rhythm.
In
1 969,
results
be of
of
effects
taken in interpreting a possible circadian
McBean
and
Halstead
(31)
reported no effect in a postprandial plasma zinc when one-third of the daily intake of zinc was taken. However, more recently Hetland and Brubakk (32) demonstrated a diurnal variation in serum zinc concentration. However, they determined zinc during only three periods, 8:00 AM, 1 1 :30 AM and 3:00 PM and they noticed a progressive fall from the first to the third specimen.
The
fall
was
not
great
but
,
rhythm
change circadian
has
not
demonstrated
a
circadian
for copper and zinc in the serum. The in serum levels do not correlate with the changes observed in the urine.
Miscellaneous
cific proteins, serum B1 2-binding capacity has been shown to be elevated by the administration of oral contraceptives (41). Recently Dawson et al. (42) were concerned as to whether some of the oral contraceptive agents contained significant concentrations of the trace minerals in the pills which might cause some of the observed changes in blood levels. Analyses, however, reveal considerable variations depending on the brands tested but as far as the trace minerals iron, copper and zinc are concerned, the highest contribution of a single pifi was 64 ;ig of iron, 2.0 pg of copper and 10.0 pg of zinc. Certainly, these are dosages which are insufficient to cause the effects noted.
amounted
to over 10% and was statistically significant. Interesting observations have been made suggesting that circadian rhythms for these trace mineral substances may be present. The first suggestion of this was by Drinker et al. (33) who showed a circadian rhythm for zinc excretion. However, more recently this phenomenon has been denied (34). Bartter (35) has also reported a circadian rhythm for copper excretion and this has been confirmed more recently by Lifschitz et al. (34). However, Lifschitz
KING
elements
The confusion regarding responses is not limited to the trace elements. Goldsmith et a!. (36, 37) and Lowenstein (38) have suggested that serum magnesium levels were lowered by contraceptive agents. These observations were made both in an epidemiological study and a controlled study in which Enovid was administered during the menstrual cycle (37). However, Thin (39) has questioned this and using a combined contraceptive agent he has demonstrated no effect on the plasma concentration of calcium or magnesium in plasma, red blood cell magnesium or platelet calcium or magnesium. In fact, he has shown that there is no change in levels of these substances during the menstrual cycle. A single report by Pulkkinen and Wiliman (40) suggests that phosphate is lowered by the contraceptive pills. Among miscellaneous effects probably related to spe-
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
research report effects of oral contraceptive the trace minerals, copper, Preliminary
on the metabolic agents on iron and zinc
The trace minerals selected for study were those where effects, at least on blood levels, had been suggested. Two of these, namely, iron and copper, have specific binding proteins. The third, zinc, does not appear to have a specific carrier protein. The initial experiment carried out was an overview of possible metabolic and/or balance changes resulting from oral contraceptive agents. A few of the results to date are summarized in this report. Experimental
design
A group of seven women, aged 18 to 32, who had been taking oral contraceptive agents for at least 2 years, volunteered as subjects. All subjects met the following criteria: nonsmoking, weight within normal range for height, free of metabolic disorders, and judged to be healthy by extensive medical histories and physical examinations. The physical characteristics of the subjects are outlined in Table 1. TABLE Subject
1 profile Age,
Code
Height,
Weight,
years
cm
kg
01
18
169.5
02
22
164.25
03
25
164.5
55.6 51.7 54.4
04 05
28 32
166.0 176.0
56.! 68.5
06
24 24
165.0 158.5
54.2 76.2
07
EFFECT The
OF ORAL
CONTRACEPTIVE
consisted of four metabolic each 28 days in length. The total length of the study was I 16 days, which included a basal period of 4 days at the beginning for subject adaptation to the routines of the metabolic unit. Oral contraceptive therapy was varied among the subjects according to the plan outlined in Table 2. During each metabolic period, oral contraceptive agents (OCA) were administered from day 6 through day 26, i.e., 21 days. OCA therapy was discontinued for 7 days before therapy was periods
initiated
study
(MP)
in the
A combined ceptive agent, and 0.1 mg steroid
next
AGENTS
ON METABOLISM
portion of the period, so the egg albumin in the basal formula was increased to provide from 0.71 to 0.74 g protein/kg in order to achieve nitrogen equilibrium during the study. Basal
corn oil
102.90 51.45
cornstarch sucrose
316.42 water 0.0002
for
use
in
this
study
and
was
biotin
0.25 MgO
estrogen-progesterone contraNorinyl (2 mg Norethindrone Mestranol) was selected as the
agent
Formula (in grams) egg albumin
39.08 81.44
cycle.
supplied to us by the Syntex Corporation in Palo Alto, California. A low dosage preparation similar to those most frequently prescribed for the public was chosen in order to study the OC effects experienced by the majority of women on steroid therapy. All women participating in the study received this preparation for at least 2 weeks prior to beginning the experiment. A controlled formula diet was fed throughout the study. Initially the following basal formula was fed in varying amounts to each subject, according to body weight so that 0.52 g protein/kg body weight was provided. This protein intake is recommended by the 1973 FAO/WHO as the safe level of protein intake for adult women. During the first period three of the six subjects were in negative nitrogen balance throughout the period and the other three were in negative nitrogen balance for a
397
1.03
MgSO4
1.36 2.53 2.64 0.90
CaCO3 KC1 K2 HPO4
The basal formula following food items:
-7H2O
NaCl
was
augmented
by
the
7.5 g beef broth in 200 g deionized water 30.0 g shortbread (Lorna Doone cookies) 25.0 g low protein rusks 37.5
g low
protein
muffins
The actual energy needs of the subjects from 33 to 37 kcal per kg. The following kilocalorie formula was used to supply tional energy. 35 28 22 13 121
ranged extra addi-
g Dextri-Maltose g cornstarch g sucrose g cornoil g.H2O
Vitamins and trace minerals other than iron, copper, and zinc were given orally in capsules. A trace mineral solution providing a quarter of the daily total of 18 mg Fe as FeSO4 15 mg Zn as ZnSO4 and 3 mg Cu as CuSO4 was fed to each subject at each meal. The total daily intake was divided into four equal feedings at 8:30 AM, 12:30 PM, 5:30 PM, and 8:30 PM. The actual daily intake of energy, nitrogen, sodium, iron and zinc per subject is shown in Table 3. ,
TABLE 2 Administration agents during
of oral the study
,
contraceptive
Metab olic period Basal Subj
(4 days)
I
II
III
IV
off
on off on
01 02
on on
on on
on on
03 04 05
on on
on off
on off
off off on
on
off
off
on
06
on
off
off0
07 o
0ffb
06 left
entered pill that
the study
the study period.
on on
on
on day
on day
14 of MP II. b 07 25 of MP II and was off the
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
Results Results to be given here are preliminary and fragmented both because of lack of time for presentation and the fact that not all analyses have been completed. I will discuss only a few of our results at this time. Blood chemistry. In order to assess health, blood was analyzed monthly for glucose, SGPT,
MARGEN
398 TABLE Energy,
3 nitrogen,
sodium,
AND
KING
iron and zinc intake Period
N, g
Na,
1
g
Periods Fe,
mg
Zn,
mg
N, g
Energy,
Subj
kcal/kg
0! 02 03 04
35 35
05 06#{176}
left
TABLE
the
mg
per day
14.8
37
6.57
2.00
19.9
14.3
35
19.5 19.7
14.8 14.8
37 37
6.15 6.42
1.94 1.97
19.7 19.9
14.3 14.3
35 33 35
5.57 6.64 5.40
1.91 2.09 1.88
19.5 19.8 19.5
14.8 14.8 14.8
37 37 37
6.60 7.88 6.39
2.01 2.21 1.98
19.7 20.1 19.7
14.3 14.3 14.3
37
8.61
2.34
19.7
14.3
study
on day
14 of MP
I.
entered
b
the study
at the beginning
ceruloplasmin indices
to
g/lOO ml
MCV,3
13.1
±
0.8
13.2
±
±
1.8
38.0
±
0.7 1.6
4.32 34.7 30.4
±
0.34
4.36
±
0.23
±
1.4
36.0
±
1.7
±
1.9
30.7
±
1.6
1.3
7.2
6.9
±
86±4 ±
2.1
urea N, uric acid and creatine. Mean values for all parameters remained normal throughout the study except SGPT which was high during metabolic periods I and II. These high means were primarily due to the values for one subject (05). See Tables 5 and 6. Hematological indices. These are shown in Table 4. No changes are seen between subjects on and off the oral contraceptives. Serum iron, TIBC, percent saturation and selected proteins. Serum iron, TIBC, percent
Kjeldah!
saturation,
LTA
transferrin,
ceruloplasmin
a2 -globu-
dure
is
oral contraceptive cycles (P 0.05). relation
to
higher
when
contraceptives
when
they
had
the not.
This
To
date
feca!
iron
and
zinc
trace
is the
procedure. associated dry solids
fecal with
increases
and
zinc
time
period
since
required
in fecal
output
also
the
for
constant.
Thus
and i.e., fecal
the
iron
that
main
analysis.
low temperature
asher
model
the
milligrams dry solids factor
Gifford-Wood Eppenbach Colloid Mill. Trace minerals not detected by our methods
S LFE
the
zinc and that
nitrogen,
increased
of the fecal output, iron or zinc per gram
relatively
3
contamination
Fecal nitrogen, iron and with fecal total weight (P < 0.00 1 ). This means
excretion
composition nitrogen,
4
element
manipulation is minimal. In the LTA procedure, oxidation is accomplished by radicals formed by high-frequency radio However, one of the disadvantages of
method
pill
treatment was found for serum iron, percent saturation, a2-globulin, or 3-globulin. Mean levels of transferrin, TIBC, percent saturation,
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
reduces
this
nificantly treatment
in
were
oral
than
method.
sample ashing oxygen waves.
were
differences
month
13-globulin
taking
analyses have been completed by atomic absorption spectroscopy. Samples were prepared for these analyses by ashing in a low temperature asher (LTA).5 This ashing proce-
lin and 3-globulin were determined three times monthly, on days 4, 11, and 18 of each metabolic cycle. Data were analyzed by the f-test to determine statistical significance of changes due to oral contraceptive treatment. Transferrin, TIBC and ceruloplasmin were sighigher during than during “off”
been
difference was not significant for serum iron or a2 -globulin. Table S summarizes these results. Fecal excretion studies. Fecal output was collected alternatively in 3- or 4-day pools in weighed plastic containers; a total of 8 collections were made each metabolic period. Each fecal pool was homogenized3 with water “free” of trace elements4 prior to laboratory analysis. Total N was analyzed by the micro-
Off
87±6
k/mm3
had
preceding
37.9
of MP III. and
subjects
On
%
significant
Zn,
19.7
RBC, k/mm’ MCHC, % MCH,zg
No
mg
1.84 1.88
Index
WBC,
Fe,
1.90
4
Hemoglobin,
g
5.55 5.20 5.42
Response of hematological oral contraceptive treatment
PCV,
Na,
Energy, kcal/kg
per day
07b
a
2, 3, and 4
LTA-604.
of
EFFECT affecting
the
apparent
OF ORAL
CONTRACEPTIVE
absorption
at any
given
On#{176} Serum
iron,
122 34
mg/lOOml
% Saturation Transferrin,
mg/100
TIBC, zg/l00
ml
ml
23
±
6
290
±
64
355
±
36 ii 0.15 0.13
Ceruloplasmin,mg/iOOml mg/lOO ml
38± 0.71 ±
3-Globulin,
0.79
6
Means
deviation. period.
mg/100
ml
are
averages
On
b
Off
contraceptive
days
during
seven of
subjects
On Pr. Mo.”
Sig.
25
NS
NS P 0.05 P 0.05 P0.0S NS
33 ± 387 ± 349 ± 46± 0.72 ±
0.12
NS
0.79
representing
contraceptive
preceding
revealed no signifior zinc apparent
and
9 40 46 13 0.15
118
various
but
no
treatment
±
±
Off Pr. Mo.d
24
60 42
contraceptive (2
Off
±
36
±
Sig.
26 8 38
N5 P 0.05 P 0.05 P 0.05 P0.05
235 ± 319 ± 43 38± 12 0.66 ± 0.12 0.72 ± 0.12
12 0.16 0.12
of
followed
cycle.
111
7
numbers
administration
to (a) above
during
preceding
±
variance in iron
and menstruation
Off1’
oral
corresponding
saturation
of
observations by
a
=
no
administered. Pr. Mo.
7-day C
NS
0.05
P ±
standard
interruption
On Pr. Mo. oral oral contraceptive
cycle.
6
Excretion
Total
=
for days
administered
administered
TABLE
21
=
#{176}
±
An analysis cant differences
percent
34 ± 244 ± 321 ± 38± 0.67 ± 0.74 ±
399
menses.
treatment
109
±
ON METABOLISM
not considered to be statistically significant due to the high variation in fecal iron and zinc. Further studies are needed to identify factors which may augment fecal output during
time appears to be the quantity of fecal output. An analysis of variance showed a statistically significant increase in fecal total weight and nitrogen excretion (P < 0.001) during menses. ( See Table 6.) Iron and zinc excretion during menses was increased also but the change was TABLES Mean values* for serum iron, iron binding capacity, selected proteins in relation to oral contraceptive
AGENTS
of nitrogen,
weight
of feces,
Dry weight Nitrogen,
iron
offeces,
and
zinc
g g
corrected
for
variations
On pill
Off
60.09 15.00
48.33 12.28
in fecal
flow
P
pill
P
Actual
Mwgen,
M.D.
and Janet
C. King,
agents
on the 2
Ph.D.
alterations noted in women taking the oral in blood levels of some trace minerals. Only a few of these, namely, iron, copper and zinc, have been studied. This paper intends to review the changes reported and outline the beginning of a series of studies to attempt to determine whether there is any metabolic significance to these alterations. The alterations reported in blood levels are generally believed to be related in large measure to alterations in levels of specific “transport” proteins induced primarily by estrogen. The significance of these changes is generally unknown, however, they involve slight elevations in serum iron and copper and possibly a reduction in serum zinc, although these are not as clear-cut as those of the other two trace minerals. In an attempt to determine whether any metabolic significance could be ABSTRACT contraceptive
Biochemical and agents have included
attributed
to
these
changes,
metabolic changes
several
studies
have
been
instituted.
The
first,
reported
in
this
paper, consisted of a metabolic study on a group of seven women aged 18-32 who had been taking oral contraceptives for at least 2 years. The subjects were studied over four metabolic periods for a total length of 1 16 days. They were studied in a random fashion, generally 2 months on and 2 months off the pill, although this varied slightly. The subjects were fed a controlled formula diet throughout this study and vitamins and trace minerals, other than copper, iron and zinc, were given orally in capsules. A trace mineral solution was administered so that the total daily iron, copper and zinc could be fed in four equal feedings per day with the formula. Measurements consisted of studies of the blood levels of the trace minerals as well as their binding proteins, various indices of blood chemistry to assess health, hematological indices and total balance studies for each of the trace minerals and nitrogen. The balance studies included not only urine and fecal loss but also skin, menstrual and hair losses. Preliminary results are discussed as well as the formidable problems involved in trace mineral studies in humans. These problems are not only related to contamination but also are analytic in nature. Am. J. Clin. Nutr. 28: 392-402, 1975.
In general, extreme paucity contraceptive lism. Most
of the zinc.
minerals, papers
contraceptive levels of alterations
can be said that there is an of data regarding the effects of
steroids on trace of the reports deal
trace The
it
agents these are
namely, report can
mineral metabowith only a few
iron,
that cause
trace mineral generally due
copper,
certain changes
the
in blood These in the
of
laneous
In
about
capacity
levels of the specific “transport” proteins. The significance, if any, of these changes is generally unknown.
(2,
More
recently,
attempts
have
been
392
The American
Journal
of Clinical
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
above
trace
substances
minerals
and
follows
1947
the
first
demonstrated that pregnancy increase on levels of iron and somewhat contraceptive elevating
made to ascertain which of the hormonal agents in the contraceptive steroids are responsible for the observed changes. There is virtually no work that I am aware of that has examined other aspects of trace mineral metabolism in relation to the contraceptive steroids. A brief review of the significant findings relating to
the related
miscel-
below.
Iron
and of
substances. to changes
each
that
(1).
These later
conclusive
observations
were
the observation pills likewise have an serum iron and iron binding
was
serum
iron
Sciences, 94720. 2This research was supported in part from the National Institutes of Health, 72-2790. 28: APRIL
1975,
that the effect on capacity
observation
the
1 Department of Nutritional California, Berkeley, California
Nutrition
followed
by
3). The important even though
pp. 392-402.
study
had a striking total iron binding
made
binding
University by a contract NIH-NICHD
Printed
in U.S.A.
of
EFFECT
capacity
OF
ORAL
CONTRACEPTIVE
is that
might be strikingly elevated in women the pill, there was no significant in hemoglobin or mean corpuscular
receiving
change hemoglobin
concentration
possible
increase
However, studies globin concentration
was,
in women receiving -the intrauterine measure
(2).
and
only
in packed cell did demonstrate on
a
whole,
slight
higher
the pill than those utilizing device as a contraceptive
In
a
comparison
between
the
effects of pregnancy and combined or sequential therapy on iron or iron binding capacity, it was found that the major increases in pregnancy occur during the second and third trimesters. Almost all of the increase was associated with an increase in transferrin (3). It was pointed out that these changes in total iron binding
capacity
(TIBC)
or
transferrin
ON
could
METABOLISM
of Norby
(7),
of a combined absorption by
volume (1). that hemo-
the
AGENTS
who
oral use of
393
studied
the influence
contraceptive radioiron.
on iron women
Six
were studied. Two absorption studies were made before oral contraceptives were administered. The first was a control and the second after a phlebotomy of 400 ml of blood. The second series of studies was similar to the first and was repeated after 3 months on oral contraceptive pills. (The removed blood was reinfused into the patients after absorption studies
were
basal
absorption
were
the
complete.)
before
same,
and
after
increased absorption was noted on and Plasma
iron
and
The
results
showed
and
during
therapy
the
of off
phlebotomy,
an
the same magnitude oral contraceptives.
TIBC
increased
during
oral
occur very rapidly, within 24 hours of institution of treatment with the contraceptive drugs.
contraceptive therapy. However, it is interesting to note that this increased serum iron and TIBC did not prevent the fall in hemoglobin after the
This
phlebotomy
an
increase increase
was
contraceptive delayed rapidly.
followed iron.
agent,
been
menstrual
the
recognized cycle
in
10 days
of the
TIBC
and
then
preliminary
some
associated
fell
contraceptives mia exhibit
that
with
the
fluctua-
tions in iron and iron binding capacities. Both the serum iron and iron binding capacity were lowest at, or immediately after, the onset of the menses when expressed as deviations from normal. Although it was observed that the mean level of iron and iron binding capacity were higher in individuals on the pill, the fluctuation within the cycle was not abolished by the administration of oral contraceptives (5).
It has
been
pointed
out
that
the
was
in
observa-
time
and
response
was
effect was from the the estrogens (4). It
for was
by
cessation
fall
from
principal rather than
immediately
After
the
7 to
for about It appeared
tions that progestogens had
not
in serum
absorption
part
no
noted.
the
change This
in
reticulocyte
finding
observation
that
who develop only a minimal
helps
women
explain
on
iron-deficiency decrease
oral
anein serum
iron. More recently pure progestogen
it has been materials
demonstrated failed to affect
(8).
that the
compound a substance is not metabolized to estrogen, was to be ineffective in changing the serum
serum levels of transferrin employed , medroxyprogesterone
which shown
The ,
levels. The
possibility
remains
separate mechanisms iron binding capacities. observed attributed
on both entirely
transferrin
that
affecting Most
there
may
iron and TIBC have to effect of the hormones
elevation.
Zilva
be
iron levels and of the changes
(9)
been on
suggests
a dis-
of iron seems to be related to the percent saturation of transferrin. The administration of oral contraceptives increased serum transferrin and TIBC to levels higher than those observed in iron-deficiency anemia. Even though iron levels are increased above normal, the percent saturation is less than that observed in the normal individual. The possibility therefore
iron binding capacity; 2) in early pregnancy there is a rise in iron levels prior to the elevation of iron binding capacity; and 3) in women on contraceptive pills the total iron binding capacity may rise without significant
exists
elevation
that
this
alteration
capacity-serum
iron
in total
relationship
iron
binding
has
attempted
to
answer
these
questions
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
must
changes
cycle
(% saturation)
may have an effect on iron transport by the gut (6). The only research I am acquainted with which
sociation
in
iron
cannot
levels
because
during
be accounted
of
these
British Medical the contraceptive
device supplying
considered
the
1)
the
menstrual
for by alterations
in
of iron.
In view the that
be
for
women.
sufficient
considerations, Journal pill
In view iron
the
editor
of
(10) has suggested may be a protective
of the to
women
problem in
of the
394
MARGEN
reproductive years, the contraceptive effect
in
conserving
nisms whereby the decrease the pill percent
iron
iron the
second
With
the
trend
with
or
one iron
of
to date
been estrogenic
of
the
potency.
In
Vrettos
et
an effect on the In pregnancy, excretion but are excretion
of
unsaturated
pregnancy
was
iron
noted.
invalidate the are primarily responsible but certainly shows that
metabolite
products
does
not
iron the
obvi-
have
a
diurnal
fluctuations
commented
that
the
early in afternoon.
on by Ramsey
plasma
iron
were
(1 2) who
reached
its
the morning and This phenomenon
the
These
stated
highest
its lowest was more
level in the recently
reinvestigated by Wiltnik et al. (13) on people eating a normal diet. They show that in the normal
free-living
variation occurring
and
in in
the
morning. reached is set
at
65-75%
plasma iron, the afternoon
lowest In
individual,
this
at
there the highest and early
night
study
is
and
the
early
lowest
values
great
values evening
in
the were
at about 4:00 AM. If the lowest value 0%, the peak values are as much as higher, reaching and gradually
4:00
PM
What
is most
interesting
these decreasing
about
levels around thereafter.
this
observation
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
be
influenced
very
markedly
were
decreased
of
Transferrin
.
transferrin
used on
crude
been well resulted
especially mestens
was total was
time to be essentially unchanged. pointed out that the method electrophoresis
It had pregnancy
with
first
in combination
as present only in pregnancy appear. Serum albumin and
gel was a relatively
capacity.
iron.
or
in Enovid) had marked of the serum proteins.
were
determination
binding
in plasma
that
workers,
in iron must be and iron
do
Proteins
should
that for this one of the
to
alone
methylester on various
out
estradiol,
haptoglobins said at that
no
correlation with the changes capacity. A word of caution into any studies of iron has
administered
(as its influence
pointed
ethinyl
varies
significant binding introduced
This
whether
et al. (14)
estrogen
by Smithies noted to
binding
to
Musa
values of the
notion
appear
as 1967 synthetic
serum it is
This
not
early the
were G0-globulin, ceruloplasmin, thyroid binding globulins and corticosteroid binding globulins. A so-called “pregnancy protein” described
whether
from 4.7 to 25 mg/24 hours. However, correlation was noted between the total binding capacity and estriol levels, although and
that
agents.
determine
estriol
As
continue effects
entirely clear are due to
urinary estrogens
The
the
effects. to
lowest
does
it is
Copper
activity
newer
still not observed
estrogenic
out, the estriol highest of all
ously estrogens effect,
Hence,
attempted.
agents, questions magnitude of
pointed are the
capacity
period.
tionships.
has
attempted
in total,
24-hour
no
However,
theory
the estrogens has binding capacity.
elevation
the
in on
it is effects
(1 1) have
over
possible effects changes of TIBC may have
some
progestational al.
transferrin
quite clear that when measuring serum iron, one must take the measurements at the same time of day and consider the dietary interrela-
to decrease
instances, the
mechaare with
of oral contraceptive to arise as to the many whether
The
brought about loss associated
absorption.
of
observed
be
KING
is that these changes in iron levels are not associated with corresponding levels in transferrin. There is less than a 10% variation in
exists that a beneficial
stores.
this may in menstrual
and the saturation
increasing test
the possibility pill may have
AND
during (15) and
It for
by
these
polyacrylyamide
technique.
known for in increase
many years that in copper levels,
the second as early as
and 1956
third
tn-
Russ and Raymunt (1 6) had demonstrated that estrogen would produce an increase in serum copper and copper
al.
binding
(17)
copper
oral
protein.
and
the
1966
in copper
sequential
or
There did excretion;
mixed
in
receiving
was
a
levels
a twofold increase 1 7- 1 8 days of the
et
increase
in subjects
There
increase
Carruthers
marked
ceruloplasmin
contraceptives.
threefold mately within
In
demonstrated
and
two-
to
approxi-
in cenuloplasmin onset of either
contraceptive
therapy.
not appear to be any effect both controls and individuals
on
urine receiv-
ing the material were said to be excreting less than 0.1 mg/liter of copper. The rise in serum concentration has been confirmed by other workers(18, 19). More
recently
Bniggs
et
al.
that progestational agents on serum copper, whereas has a marked effect on increasing shown effect
In fact,
the
increase
in serum
(18,
20)
have
alone have no estrogen alone serum copper.
copper
appears
to
EFFECT
OF ORAL
CONTRACEPTIVE
be proportional to the dose of estrogen administered. It is rather interesting to note that the serum copper and ceruloplasmin levels in pregnancy tend to reach their maximum at about the 22nd to 25th week of gestation and then gradually fall to the time of delivery (21). More recently, Daunter and Elstein (22) have evaluated the levels of serum copper, ceruloplasmin
and
ceruloplasmin
oxidase
activity
in
women using copper-containing intrauterine devices and in women taking combined oral contraceptives. No effect on serum copper or ceruloplasmin levels was noted in women wearing the intrauterine device. The levels reported although
were well no normal
within controls
study. The women agents were of two 1 .0 mg Norethisterone
receiving groups. plus
tradiol
and
Norethisterone ethinylestradiol. ceruloplasmin tivity were suggests
that
the
other,
half
the normal were used
oral contraceptive One group received SO zg of ethinylesthe
with the same No differences and ceruloplasmin noted in these two the
effect
was
limits, in this
almost
amount
of
amount of in copper, oxidase acgroups. This entirely
due
to the estrogen. Schenker et al. (23) have also studied a group of women taking the oral contraceptive agents with and without choleostasis. The oral contraceptive agents demonstrate
the
usual
increase
in copper
and
cerulo-
plasmin. Also, correlation coefficients between copper and ceruloplasmin is extremely high, i.e., increases in the patients receiving oral contraceptives (r = 0.88) versus the controls (r = 0.74). Patients who develop choleostasis while on the oral contraceptive agents have even higher copper and ceruloplasmin, with very high correlation (r 0.94) when this condition is superimposed. Olatunbosun
et
al.
(24)
have
studied
the
effect of oral contraceptives on serum copper levels in Nigerian women. In this study they were interested in the suggestion that thromboembolic phenomena are rarer in black United States females than in white (25). They studied patients with the intrauterine device in place versus women on two different oral contraceptive agents. As noted in prior studies, the intrauterine device had no effect on serum copper levels although the oral contraceptive agents produced the expected rise in copper. However, in general, the degree of elevation was
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
AGENTS lower
in
ON METABOLISM this
been 30%. The reason for this is unclear, although as pointed out, women in Nigeria frequently have lower serum albumin and also frequently reported,
group
i.e.,
elevated
an
than
395
increase
7-globulin.
had
of
Whether
usually
only
about
these
lesser
eleva-
tions are due to changes in serum protein levels due to genetic or nutritional differences, is also quite unclear. From a metabolic point of view, the significance of these changes in copper and ceruloplasmin levels are totally unknown and quite confusing. This is particularly true if one accepts the suggestion that ceruloplasmin is not considered to be a copper transport protein in the manner of transferrin (26). In fact, it is suggested that the metabolically active and exchangeable copper, i.e., that which is derived from dietary sources and other free copper, is principally transported by albumin and amino acids.
The
latter
copper
is in equilibrium
with
a
pool of “free” copper in the erythrocytes (27). More recently, a number of copper-binding proteins have been described but their activity, as well as the general action of the free or poorly bound copper, is largely a mystery at this time. It appears that one of the major functions of ceruloplasmin may be related to the release of bound iron but what regulates changes in the levels of this circulating enzyme, is unknown. However, the possibility exists that in pregnancy, when there is a great demand for transfer
of
iron,
copper
and
across the placental membrane, these transport systems may induced.
other
materials
mobilization be hormonally
of
Zinc
Several observations have been made regarding zinc levels and to date there does not appear to be a universal agreement as to the effects of the steroids on this mineral. Halstead in 1968 (28)
stated
that
both
pregnancy
and
oral
contraceptive agents levels although he
cause a fall in serum zinc confirmed the elevation
previously
in
observed
copper
levels
in
these
two conditions. In the following year, O’Leary and Spellacy (29) found that although there seemed to be a slightly lowered zinc level in pregnancy this was not statistically significant, although the oral contraceptive agents which he employed markedly elevated the zinc levels. More
recently
these
observations
of
O’Leary
396
MARGEN
AND
have been confirmed by reported measurements of Schenker et al. (30). The above observations point out the care which must be used
in
interpreting
of
the
reported
these materials. Again, care must zinc levels because rhythm.
In
1 969,
results
be of
of
effects
taken in interpreting a possible circadian
McBean
and
Halstead
(31)
reported no effect in a postprandial plasma zinc when one-third of the daily intake of zinc was taken. However, more recently Hetland and Brubakk (32) demonstrated a diurnal variation in serum zinc concentration. However, they determined zinc during only three periods, 8:00 AM, 1 1 :30 AM and 3:00 PM and they noticed a progressive fall from the first to the third specimen.
The
fall
was
not
great
but
,
rhythm
change circadian
has
not
demonstrated
a
circadian
for copper and zinc in the serum. The in serum levels do not correlate with the changes observed in the urine.
Miscellaneous
cific proteins, serum B1 2-binding capacity has been shown to be elevated by the administration of oral contraceptives (41). Recently Dawson et al. (42) were concerned as to whether some of the oral contraceptive agents contained significant concentrations of the trace minerals in the pills which might cause some of the observed changes in blood levels. Analyses, however, reveal considerable variations depending on the brands tested but as far as the trace minerals iron, copper and zinc are concerned, the highest contribution of a single pifi was 64 ;ig of iron, 2.0 pg of copper and 10.0 pg of zinc. Certainly, these are dosages which are insufficient to cause the effects noted.
amounted
to over 10% and was statistically significant. Interesting observations have been made suggesting that circadian rhythms for these trace mineral substances may be present. The first suggestion of this was by Drinker et al. (33) who showed a circadian rhythm for zinc excretion. However, more recently this phenomenon has been denied (34). Bartter (35) has also reported a circadian rhythm for copper excretion and this has been confirmed more recently by Lifschitz et al. (34). However, Lifschitz
KING
elements
The confusion regarding responses is not limited to the trace elements. Goldsmith et a!. (36, 37) and Lowenstein (38) have suggested that serum magnesium levels were lowered by contraceptive agents. These observations were made both in an epidemiological study and a controlled study in which Enovid was administered during the menstrual cycle (37). However, Thin (39) has questioned this and using a combined contraceptive agent he has demonstrated no effect on the plasma concentration of calcium or magnesium in plasma, red blood cell magnesium or platelet calcium or magnesium. In fact, he has shown that there is no change in levels of these substances during the menstrual cycle. A single report by Pulkkinen and Wiliman (40) suggests that phosphate is lowered by the contraceptive pills. Among miscellaneous effects probably related to spe-
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
research report effects of oral contraceptive the trace minerals, copper, Preliminary
on the metabolic agents on iron and zinc
The trace minerals selected for study were those where effects, at least on blood levels, had been suggested. Two of these, namely, iron and copper, have specific binding proteins. The third, zinc, does not appear to have a specific carrier protein. The initial experiment carried out was an overview of possible metabolic and/or balance changes resulting from oral contraceptive agents. A few of the results to date are summarized in this report. Experimental
design
A group of seven women, aged 18 to 32, who had been taking oral contraceptive agents for at least 2 years, volunteered as subjects. All subjects met the following criteria: nonsmoking, weight within normal range for height, free of metabolic disorders, and judged to be healthy by extensive medical histories and physical examinations. The physical characteristics of the subjects are outlined in Table 1. TABLE Subject
1 profile Age,
Code
Height,
Weight,
years
cm
kg
01
18
169.5
02
22
164.25
03
25
164.5
55.6 51.7 54.4
04 05
28 32
166.0 176.0
56.! 68.5
06
24 24
165.0 158.5
54.2 76.2
07
EFFECT The
OF ORAL
CONTRACEPTIVE
consisted of four metabolic each 28 days in length. The total length of the study was I 16 days, which included a basal period of 4 days at the beginning for subject adaptation to the routines of the metabolic unit. Oral contraceptive therapy was varied among the subjects according to the plan outlined in Table 2. During each metabolic period, oral contraceptive agents (OCA) were administered from day 6 through day 26, i.e., 21 days. OCA therapy was discontinued for 7 days before therapy was periods
initiated
study
(MP)
in the
A combined ceptive agent, and 0.1 mg steroid
next
AGENTS
ON METABOLISM
portion of the period, so the egg albumin in the basal formula was increased to provide from 0.71 to 0.74 g protein/kg in order to achieve nitrogen equilibrium during the study. Basal
corn oil
102.90 51.45
cornstarch sucrose
316.42 water 0.0002
for
use
in
this
study
and
was
biotin
0.25 MgO
estrogen-progesterone contraNorinyl (2 mg Norethindrone Mestranol) was selected as the
agent
Formula (in grams) egg albumin
39.08 81.44
cycle.
supplied to us by the Syntex Corporation in Palo Alto, California. A low dosage preparation similar to those most frequently prescribed for the public was chosen in order to study the OC effects experienced by the majority of women on steroid therapy. All women participating in the study received this preparation for at least 2 weeks prior to beginning the experiment. A controlled formula diet was fed throughout the study. Initially the following basal formula was fed in varying amounts to each subject, according to body weight so that 0.52 g protein/kg body weight was provided. This protein intake is recommended by the 1973 FAO/WHO as the safe level of protein intake for adult women. During the first period three of the six subjects were in negative nitrogen balance throughout the period and the other three were in negative nitrogen balance for a
397
1.03
MgSO4
1.36 2.53 2.64 0.90
CaCO3 KC1 K2 HPO4
The basal formula following food items:
-7H2O
NaCl
was
augmented
by
the
7.5 g beef broth in 200 g deionized water 30.0 g shortbread (Lorna Doone cookies) 25.0 g low protein rusks 37.5
g low
protein
muffins
The actual energy needs of the subjects from 33 to 37 kcal per kg. The following kilocalorie formula was used to supply tional energy. 35 28 22 13 121
ranged extra addi-
g Dextri-Maltose g cornstarch g sucrose g cornoil g.H2O
Vitamins and trace minerals other than iron, copper, and zinc were given orally in capsules. A trace mineral solution providing a quarter of the daily total of 18 mg Fe as FeSO4 15 mg Zn as ZnSO4 and 3 mg Cu as CuSO4 was fed to each subject at each meal. The total daily intake was divided into four equal feedings at 8:30 AM, 12:30 PM, 5:30 PM, and 8:30 PM. The actual daily intake of energy, nitrogen, sodium, iron and zinc per subject is shown in Table 3. ,
TABLE 2 Administration agents during
of oral the study
,
contraceptive
Metab olic period Basal Subj
(4 days)
I
II
III
IV
off
on off on
01 02
on on
on on
on on
03 04 05
on on
on off
on off
off off on
on
off
off
on
06
on
off
off0
07 o
0ffb
06 left
entered pill that
the study
the study period.
on on
on
on day
on day
14 of MP II. b 07 25 of MP II and was off the
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
Results Results to be given here are preliminary and fragmented both because of lack of time for presentation and the fact that not all analyses have been completed. I will discuss only a few of our results at this time. Blood chemistry. In order to assess health, blood was analyzed monthly for glucose, SGPT,
MARGEN
398 TABLE Energy,
3 nitrogen,
sodium,
AND
KING
iron and zinc intake Period
N, g
Na,
1
g
Periods Fe,
mg
Zn,
mg
N, g
Energy,
Subj
kcal/kg
0! 02 03 04
35 35
05 06#{176}
left
TABLE
the
mg
per day
14.8
37
6.57
2.00
19.9
14.3
35
19.5 19.7
14.8 14.8
37 37
6.15 6.42
1.94 1.97
19.7 19.9
14.3 14.3
35 33 35
5.57 6.64 5.40
1.91 2.09 1.88
19.5 19.8 19.5
14.8 14.8 14.8
37 37 37
6.60 7.88 6.39
2.01 2.21 1.98
19.7 20.1 19.7
14.3 14.3 14.3
37
8.61
2.34
19.7
14.3
study
on day
14 of MP
I.
entered
b
the study
at the beginning
ceruloplasmin indices
to
g/lOO ml
MCV,3
13.1
±
0.8
13.2
±
±
1.8
38.0
±
0.7 1.6
4.32 34.7 30.4
±
0.34
4.36
±
0.23
±
1.4
36.0
±
1.7
±
1.9
30.7
±
1.6
1.3
7.2
6.9
±
86±4 ±
2.1
urea N, uric acid and creatine. Mean values for all parameters remained normal throughout the study except SGPT which was high during metabolic periods I and II. These high means were primarily due to the values for one subject (05). See Tables 5 and 6. Hematological indices. These are shown in Table 4. No changes are seen between subjects on and off the oral contraceptives. Serum iron, TIBC, percent saturation and selected proteins. Serum iron, TIBC, percent
Kjeldah!
saturation,
LTA
transferrin,
ceruloplasmin
a2 -globu-
dure
is
oral contraceptive cycles (P 0.05). relation
to
higher
when
contraceptives
when
they
had
the not.
This
To
date
feca!
iron
and
zinc
trace
is the
procedure. associated dry solids
fecal with
increases
and
zinc
time
period
since
required
in fecal
output
also
the
for
constant.
Thus
and i.e., fecal
the
iron
that
main
analysis.
low temperature
asher
model
the
milligrams dry solids factor
Gifford-Wood Eppenbach Colloid Mill. Trace minerals not detected by our methods
S LFE
the
zinc and that
nitrogen,
increased
of the fecal output, iron or zinc per gram
relatively
3
contamination
Fecal nitrogen, iron and with fecal total weight (P < 0.00 1 ). This means
excretion
composition nitrogen,
4
element
manipulation is minimal. In the LTA procedure, oxidation is accomplished by radicals formed by high-frequency radio However, one of the disadvantages of
method
pill
treatment was found for serum iron, percent saturation, a2-globulin, or 3-globulin. Mean levels of transferrin, TIBC, percent saturation,
Downloaded from https://academic.oup.com/ajcn/article-abstract/28/4/392/4716472 by Washington University in St. Louis user on 09 April 2018
reduces
this
nificantly treatment
in
were
oral
than
method.
sample ashing oxygen waves.
were
differences
month
13-globulin
taking
analyses have been completed by atomic absorption spectroscopy. Samples were prepared for these analyses by ashing in a low temperature asher (LTA).5 This ashing proce-
lin and 3-globulin were determined three times monthly, on days 4, 11, and 18 of each metabolic cycle. Data were analyzed by the f-test to determine statistical significance of changes due to oral contraceptive treatment. Transferrin, TIBC and ceruloplasmin were sighigher during than during “off”
been
difference was not significant for serum iron or a2 -globulin. Table S summarizes these results. Fecal excretion studies. Fecal output was collected alternatively in 3- or 4-day pools in weighed plastic containers; a total of 8 collections were made each metabolic period. Each fecal pool was homogenized3 with water “free” of trace elements4 prior to laboratory analysis. Total N was analyzed by the micro-
Off
87±6
k/mm3
had
preceding
37.9
of MP III. and
subjects
On
%
significant
Zn,
19.7
RBC, k/mm’ MCHC, % MCH,zg
No
mg
1.84 1.88
Index
WBC,
Fe,
1.90
4
Hemoglobin,
g
5.55 5.20 5.42
Response of hematological oral contraceptive treatment
PCV,
Na,
Energy, kcal/kg
per day
07b
a
2, 3, and 4
LTA-604.
of
EFFECT affecting
the
apparent
OF ORAL
CONTRACEPTIVE
absorption
at any
given
On#{176} Serum
iron,
122 34
mg/lOOml
% Saturation Transferrin,
mg/100
TIBC, zg/l00
ml
ml
23
±
6
290
±
64
355
±
36 ii 0.15 0.13
Ceruloplasmin,mg/iOOml mg/lOO ml
38± 0.71 ±
3-Globulin,
0.79
6
Means
deviation. period.
mg/100
ml
are
averages
On
b
Off
contraceptive
days
during
seven of
subjects
On Pr. Mo.”
Sig.
25
NS
NS P 0.05 P 0.05 P0.0S NS
33 ± 387 ± 349 ± 46± 0.72 ±
0.12
NS
0.79
representing
contraceptive
preceding
revealed no signifior zinc apparent
and
9 40 46 13 0.15
118
various
but
no
treatment
±
±
Off Pr. Mo.d
24
60 42
contraceptive (2
Off
±
36
±
Sig.
26 8 38
N5 P 0.05 P 0.05 P 0.05 P0.05
235 ± 319 ± 43 38± 12 0.66 ± 0.12 0.72 ± 0.12
12 0.16 0.12
of
followed
cycle.
111
7
numbers
administration
to (a) above
during
preceding
±
variance in iron
and menstruation
Off1’
oral
corresponding
saturation
of
observations by
a
=
no
administered. Pr. Mo.
7-day C
NS
0.05
P ±
standard
interruption
On Pr. Mo. oral oral contraceptive
cycle.
6
Excretion
Total
=
for days
administered
administered
TABLE
21
=
#{176}
±
An analysis cant differences
percent
34 ± 244 ± 321 ± 38± 0.67 ± 0.74 ±
399
menses.
treatment
109
±
ON METABOLISM
not considered to be statistically significant due to the high variation in fecal iron and zinc. Further studies are needed to identify factors which may augment fecal output during
time appears to be the quantity of fecal output. An analysis of variance showed a statistically significant increase in fecal total weight and nitrogen excretion (P < 0.001) during menses. ( See Table 6.) Iron and zinc excretion during menses was increased also but the change was TABLES Mean values* for serum iron, iron binding capacity, selected proteins in relation to oral contraceptive
AGENTS
of nitrogen,
weight
of feces,
Dry weight Nitrogen,
iron
offeces,
and
zinc
g g
corrected
for
variations
On pill
Off
60.09 15.00
48.33 12.28
in fecal
flow
P
pill
P
Actual
