The Journal of the Pakistan Medical Association

Rawalpindi - IslamabadISSN-0303-5212
Volume 32, Number 2, Jul - Dec 2007
Original article
Relationship between maternal hemoglobin and Perinatal outcome
Umber Jalil Bakhtiar, Yasmeen Khan, Razia Nasar
Correspondence:
From Department of Obstetrics and Gynecology
Islamic International Medical College Trust,
Railway Hospital, Rawalpindi.
Received: February 8, 2007 Accepted: June 18, 2007
ABSTRACT
Objective: To Study the Relationship between Maternal Hemoglobin and Perinatal
outcome in a cohort of 860 pregnant women and to highlight the importance of antenatal
care regarding maternal health and fetal outcome.
Methods: All Singleton pregnancies delivering at Pakistan Railway Hospital Rawalpindi
from January 2004 to December 2005 that fulfilled the required criteria were included.
Results: Out of the 860 patients, 402 were anemic (<11gm/dl) and 458 were non anemic.
Perinatal outcome included preterm delivery, low birth weight, intrauterine growth
retardation, perinatal death, low apgr scores and intrauterine fetal deaths. Risk of preterm
and Low birth weight among anemic women was 3.4 and 1.8 times more than non
anaemic women. The neonates of anemic woman also had 1.7 times increased risk of
having low Apgar scores at 1 min. Among anemic women there was 2.2 times greater risk
of intrauterine fetal death than the non-anemic women.
Conclusions: Regular antenatal care from first trimester has a vital role in assessing and
managing maternal anemia timely and it directly affects the perinatal outcome. The
patients with anemia have also higher risk of having low birth weight, preterm births and
intra uterine fetal death. (Rawal Med J 2007;32:102-104).
Maternal hemoglobin, perinatal outcome, preterm birth
INTRODUCTION
Maternal anemia is a common problem in pregnancy in developing countries like
Pakistan. According to WHO anemia is defined as the hemoglobin level of <11 gm /dl.
Anemia due to Iron Deficiency has serious health and functional consequences and the
most of its nutritional component is controllable with a very high benefit/cost ratio.1 It is
estimated that 1,200 million people are anemic globally.2 The prevalence of anemia
depends upon socio economic status lifestyle, parity, associated medical problems and
regular antenatal care. Maternal anemia in pregnancy is commonly considered as risk
factor for poor pregnancy out come and can threaten the life of mother and fetus.3
However, the extent to which the maternal hemoglobin concentration affects the fetal
weight and fetal out come is still uncertain. Some studies have shown a strong association

between low hemoglobin before delivery and adverse out come4 while other studies have
not found a significant association.5 Thus the aim of this study was to evaluate the
antenatal maternal hemoglobin and find its impact on perinatal outcome.
MATERIALS AND METHODS
The study was carried out on pregnant women attending Obstetric outpatient Department
of Pakistan Railway Hospital Rawalpindi from January 2004 to December 2005. Total
number of women delivering during this period was 1578. Women attending the out
clinic before 24 were of gestation, with singleton pregnancy and ages 18 and above were
included in the study. Women with multiple pregnancies, past preterm labor and other
associated medical complications were excluded. Anemia was defined according to WHO
criteria i.e. Haemoglobin <11 gm/dl .In this study, anemia was labeled if Hb was
<11gm/dl on 2 occasions during pregnancy and labour.
Hemoglobin levels were measured at first visit, then at the end of second trimester and
twice in the third trimester. If the Hemoglobin levels were <9 gm/dl after 34 weeks than
parenteral iron therapy either intramuscularly or intravenously were given. At Hb
<7gm/dl blood transfusions were given. All the information regarding gestational age at
delivery, complications at delivery, fetal outcome in term of weight and Apgar score was
recorded.
RESULTS
A total of 860 women fulfilled the inclusion criteria, 420 in non anemic group and 540 in
anemic group. Twenty-one women were lost to follow-up after first interview.
Mean age of the women in anemic group was 25.85 and 24.20 in non-anemic group.
There was no statistically significant difference between the two groups in terms of
education level up to primary (table 1).
Table 1. Demographic characteristics of two groups. (n=860)

Risk of preterm deliveries (<37 wks) was 3.4 times greater in anemic women than the
non-anemic. There were 1.8 and 1.7 times greater risk of Low birth rate and Intra uterine
growth retarded babies in anemic group (table 2). Perinatal mortality was 3.5 times more
in anemic group. Low Apgar at 1 min and intra uterine fetal death also showed increased
ratio in this group.
Table 2. Perinatal Outcome among the anemic and non-anemic women.

DISCUSSION
In the developing world, current strategies, to prevent and correct anemia and iron
deficiency in pregnant women have met with little success.6 Two large studies in the
industrial world, involving over one million pregnancies clearly indicated that favorable
pregnancy outcomes are less frequent among anemic mothers.7,8 The causality of anemia
in these undesirable pregnancy outcomes has been established further by studies that
show the positive results obtained in births weights and perinatal deaths by the successful
treatment of anemia with iron and folic acid as low birth weight (<2000g) was reduced

from 50% to 7% and perinatal mortality from 38% to 4% in a study in Nigeria.9

Our data showed association of maternal anemia in pregnancy with increased risk of
delivery of premature and LBW babies, intrauterine death and low Apgar score at one
minute. These deaths were commonly found to be due to prematurity and sepsis. Since
our study was done at a tertiary care unit and majority of women in Pakistan deliver at
home, it is expected that the burden of anemia and its effect on pregnancy outcome is
much greater outside the hospital settings. Maternal hemoglobin values during pregnancy
are associated with LBW and preterm birth in a U-shaped relationship with a high rate of
low birth weight at low and high concentration of maternal hemoglobin.10
It is estimated that 7.3 million perinatal deaths occur annually in the world 11 and, by
correcting anemia; many of these deaths can be prevented. This study also showed high
ratio of perinatal deaths due to prematurity. Anemia with <Hb 8gm/dl has seem to be
associated with birth weight values that are 200-300g lower than in women with >10g/l
hemoglobin.12 The relationship between anemia and infection has also been proposed13
as Corticotrophin Releasing Hormones plays a role in causing preterm labor or premature
rupture of membranes.14 Presentation for antenatal care in third trimester is common in
our country.15 Since demand for micronutrients is maximum in third trimester, this could
be one factor underlying the high prevalence of anemia.16 In conclusion, this study
showed an association of maternal anemia with increased prematurity, LBW, intra uterine
deaths and low Apgar scores. Other nutritional deficiencies can also be the causative
factors. Further studies are needed in this area.
REFERENCES
Viteri FE. The consequences of iron deficiency and anemia in pregnancy. In; Nutrient
Regulation during Pregnancy, Lactation and infant Growth. L. Alien, J, King and
B.Lonnerdal. Eds. 1994 Plenom Press, New York, PP. 121-133.
WHO. National. Strategies for overcoming Micronutrient Malnutrition. Document EB
1991; 89/27. Executive Board, 89th Session.
Gregory P, Taslim A. Health Status of the Pakistani population: a health profile and
comparison with the United States. Am J Public Health 2001; 91:93-8
Iron deficiency anemia: Re-examining the nature and magnitude of the public health
problem. Proceedings of a conference. May 21-24, 2000. Belmont, Maryland, USA. J
Nutrition 2001;131:5635-7038.
Karim SA et al. Anemia in pregnancy-its cause in the underprivileged class of Karachi. J
Pak Med Assoc 1994;44:90-2..
ACC/SCN (1991) Controlling bar Deficiency. A report based on an ACC/SCN workshop.
S.Gillespie, J. Kevany and J. Mason, eds. ACC/SCN State of the Art series. Nutrition
policy discussion paper No. 9 ACC/SCN C/O WHO, Cieneva, Switzerland.
Garn SM., Ridella SA, Petzold AS, Falkner F. Maternal Hematological level and
pregnancy Outcomes. Sem. In Perinatop 1981;5:115-162.
Murphy JF, New combe RG, Coles EG, Pearson JF. Relation of Hemoglobin levels in
First and Second Trimesters to Outcome of Pregnancy. Lancet 1986; I:992-995.
Fleming AF. A study of Anemia of Pregnancy in Ibadan, Western Nigeria with special
Reference to Folic acid Deficiency. MD Thesis, University of Cambridge. 1991 Quoted
by A. Hughes in Anema of Pregnancy Maternal Health and Safe Motherhood, WHO,
1991.

Verma KC, Dhar G. Relationship of maternal anemia, birth weight and perinatal
mortality: a hospital study. Indian Pediatr 1976;13:469-41.
Shazia T, Faheem S, Saad R. Perinatal mortality: A survey. Pak J Obstet Gynaecol
1994;7:1-8.
Steer P, Alam MA, Wadsworh J, Welch A. Relationship between maternal haemoglobin is
concentration and birth weight in different ethnic groups. Brit Med J 1995;310:489-91.
Hooton TM, Scholes D, Hughes JP, Winter C, Roberts PL, Stapleton AE, et al. A
prospective study of risk factors for symptomatic urinary tract infection in young women.
N Engl J Med 1996; 335:468-74.
Khan MM. Effect of maternal anemia on fetal parameters. J Ayub Med Coll Abbottabad
2001;13:38-41.
Islam MZ, Lamberg AC, Bhuyan MA, Salamatulla Q. Iron status of premenopausal
women in two regions of Bangladesh: Prevalence of deficiency in high and low socioeconomic group. Eur J Clin Nutr 2001;55: 598-604.
Blot DD, Tehernin G. Iron deficiency in pregnancy: effects on the newborn. Curr Opin
Hematol 1999;6:65-70.

EFFECT OF MATERNAL ANEMIA AT HIGH ALTITUDE ON INFANT

HEMATOCRIT AND OXYGENATION
MARIA E. RAMIREZ-CARDICH, MAYUKO SAITO, ROBERT H. GILMAN, LUIS E.
ESCATE, JOHN J. STROUSE, CHRISTOPHER KABRHEL, COLLEEN JOHNSON,
RIVKA GALCHEN, AND CHRISTIAN T. BAUTISTA
Asociacin Benfica Proyectos en Informtica, Salud, Medicina y Agricultura (A. B.
PRISMA), Lima, Peru; Johns Hopkins Bloomberg School of Public Health, Baltimore,
Maryland; Department of Pathology, Universidad Peruana Cayetano Heredia, School of
Public Health, Health Administration and School of Medicine, Lima, Peru; EsSalud
Hospital, La Oroya, Peru; Department of Pediatrics, Johns Hopkins University School of
Medicine, Baltimore, Maryland; Mount Sinai School of Medicine, New York, New York

 ABSTRACT
Hematocrit levels were determined in 36 mothers living at high altitudes (3,750 meters)
and their infant cord bloods to determine the effect of maternal anemia on the infant. The
arterial oxygen saturation (SaO2) and respiratory rate of the infants were also followed
during the first four months of life. There was a negative correlation between maternal
hematocrit and infant hematocrit (rs = 0.57). Nineteen babies born to anemic women
(hematocrit < 41%) had a significantly higher mean hematocrit (59.9%) than those born
to non-anemic mothers (55.8%; P = 0.003). The SaO2 levels and respiratory rates of
infants were not different between infants born to non-anemic and anemic mothers. At
high altitudes, infants from mothers with anemia have higher hematocrits than those born
to non-anemic mothers.
INTRODUCTION
Fetal oxygenation is critically dependent on maternal oxygen supply. Findings of reduced
infant birth weight and increased neonatal and infant mortality suggest that fetuses at
high altitudes experience greater hypoxia than those at low altitudes.14 However,
infants born to Tibetan women native to the highlands had a higher birth weight than
those of women native to lower altitudes, suggesting that genetic or physiologic
adaptation may permit adequate oxygenation of fetuses at high altitudes.58
In animal studies, maternal systemic and uteroplacental circulations are capable of
adapting well to chronic maternal anemia, even in the case of acute anemia.9,10
Studies at high altitudes are useful in evaluating the effect of maternal anemia on infant
hematocrit values and oxygenation because the hypoxia of high altitude produces an
exaggerated response compared with those seen at sea level. It must be emphasized that
to date no studies have determined the effect in humans of maternal anemia on newborn
oxygenation at high altitudes. In fact, the effect of maternal anemia in the newborn, even
at sea level, is an inexplicably neglected area. We examined mothers hematocrit levels
and compared them to those of their infants, and also observed arterial oxygen saturation
(SaO2) at birth. In addition, we monitored arterial oxygen saturation of the infants born at
high altitudes during the first four months of life to determine the SaO2 response of
infants over time and with varying states of activity (awake, sleeping, and feeding) to
evaluate the relationship between maternal anemia and neonatal oxygenation.
METHODS
Site. The study site was the EsSalud Hospital in La Oroya, Peru (altitude = 3,750 meters
[12,300 feet]). This hospital only admits patients with social security insurance. These
individuals usually have stable employment (miners, road workers, and employees) and
are somewhat more affluent than the surrounding general population. They come from
nearby towns that are at an altitude similar to La Oroya (between 3,500 and 4,000
meters).
Subjects. Women were eligible for the study if they were between 16 and 40 years of age,
had normal results on a prenatal examination, and were expected to have a normal

vaginal delivery. Patients were excluded for any of the following pre-delivery
complications: pre-eclampsia, eclampsia (the occurrence in a woman with preeclampsia
of seizures that cannot be attributed to other causes), maternal dystocia, placenta previa,
or severe hemorrhage during the birth process. All women were of mixed (Mestizo)
descent (Quechua native and Spanish). All were born at high altitudes (more than 3,000
meters) and lived at altitudes ranging from 3,500 to 4,000 meters throughout their lives.
All deliveries occurred in the hospital. An experienced pediatrician evaluated all
newborns, and only term infants (more than 37 weeks based on expected date of
confinement) with a normal cardiopulmonary examination result and an Apgar score of
eight or more were eligible for the study. Gestational age was determined by the method
of Capurro and others using five physical examinations and two neurologic signs.11
Low birth weight infants (< 2,500 grams) were excluded from the analysis.
After informed consent was obtained from the pregnant woman, a maternal history was
taken that included place of residence, birthplace, and use of alcohol, drugs, or tobacco.
The study protocol was reviewed and approved by the ethical committee at A.B.
PRISMA (Lima, Peru) and the institutional review board of the Johns Hopkins
Bloomberg School of Public Health (Baltimore, MD).
Study techniques. The mothers hematocrit was measured within 12 hours prior to
delivery, and the newborn hematocrit was measured in cord blood immediately postdelivery. The hematocrit levels were determined by the microcapillary method.
Prevalence of anemia was estimated in pregnant women after correcting for the plasma
volume expansion that occurs during pregnancy. Persons who fall below a set level of the
fifth percentile of the non-anemic population, assuming a normal distribution, are
considered to be anemic. We calculated the hemoglobin (Hb) cut-off for women of
childbearing age at 3,750 meters using the model developed by Cohen and Haas12
according to the following formula: Hb cut-off = mean Hb [(0.061 x mean Hb) x1.96]
when the mean Hb (g/L) = 120 + 16.3 x [exp (0.00038 x (altitude 1000))].
This model demonstrates the change in hematocrit of non-anemic women at different
altitudes and predicts that at an altitude of 3,750 meters women of childbearing age have
Hb cutoffs for anemia of 14.6 g/dL (hematocrit = 44%). However, the degree of
hemodilution that occurs in pregnant woman at high altitudes has not been defined. At
this altitude, pregnant women in the second and third trimester were estimated to have an
Hb level of 1 g/dL less than their level in the non-pregnant state.12 This gives an Hb
cut-off for anemia of 13.6g/dL (hematocrit = 41%). This hematocrit level of 41% was
used for the comparison between high altitude anemic and non-anemic mothers.
Arterial oxygen saturation (SaO2) was measured by pulse oximetry with an N20 pulse
oximeter (Nellcor, Inc., Pleasanton, CA). An appropriately sized sensor was placed on the
lateral foot to minimize postductal effects.13 None of the infants received
supplemental oxygen at delivery or afterwards. Pulse oximetry studies were performed at
varied activity levels at 10 minutes to 4 months of age (10 minutes, 2 hours, 6 hours, 12
hours, 24 hours, 1 week, 2 weeks, 3 weeks, 4 weeks, 2 months, 3 months, and 4 months)
after birth. Each child was studied, if possible, while quietly awake, while feeding, and
while quietly sleeping.14 A child was considered quietly awake when his eyes were
open and he or she was not crying. Feeding was defined as actively sucking and
swallowing from a breast or a bottle. Quietly asleep was defined by the clinical criteria of
Prechtl:15 a relaxed state with eyes closed, a still, regular respiration and pulse rate,

and no gross movements except occasional startles. The infants were observed for three
minutes in a definable state before collecting data. The SaO2 levels were recorded at oneminute intervals for a total of 10 minutes during each of the three activities.6,13
Care was taken to ensure that the pulse rate displayed on the oximeter reflected the heart
rate heard by stethoscope. Respiratory rate was counted by auscultation for a one-minute
period during quiet sleep and the awake state only. Data was rejected when the infants
activity was not clearly classifiable or when the pulse oximetry was less than 65% (two
children on one occasion) because pulse oximetry readings of SaO2 less than 65% have
been shown to be inaccurate.16 Infants were screened for respiratory illness before
each home study; if fever or signs of lower respiratory infection were present, they were
excluded from further study (one child on one occasion). Infants with nasal congestion
alone were not excluded because in previous studies at high altitudes we demonstrated
that children with upper respiratory infections had SaO2 readings similar to those of
normal children.17
Statistical analysis. The chi-square or Fishers exact tests were used to compare
differences in proportions. The Student t-test or Mann-Whitney U test was used to
compare continuous variables. To evaluate a linear relationship between maternal and
cord blood hematocrit, a Spearmans correlation coefficient (rs) was calculated. Tworelated measures were analyzed using the Wilcoxon signed rank test. Analysis of slopes
was performed for comparison of arterial oxygen saturation and respiratory rate for
awake infants at different time intervals by the general linear models for repeated
measures analysis in anemic and non-anemic women. All reported confidence intervals
are 95% and all reported P values are two-sided. Statistical analyses were carried out
using SPSS for Windows version 10.1 (SPSS Corporation, Chicago, IL).
RESULTS
The 36 mother-infant pairs included in the study remained in the area for the study
period. None of the mothers or infants reported using medications, drinking alcohol, or
smoking during their pregnancies or needed supplemental oxygen at any time. None of
the mothers had meconium-stained fluid at delivery. The mean gestational age was 38.9
weeks (Standard error of the mean [SEM] = 0.2) by the method of Capurro and others.
11 Of the 36 women, 19 (53%) were anemic with a mean hematocrit of 34.9% (SEM =
1.1) compared with 43.4% (SEM = 0.3) for the 17 non-anemic women (Table 1 ).

The newborns had a mean umbilical cord venous blood hematocrit of 57.9% (SEM = 0.7)
(Table 1 ). There was a moderate negative correlation between maternal and cord
blood hematocrit (rs = 0.57, P < 0.0001) (Figure 1 ). The infants born to anemic
mothers had a significantly higher mean hematocrit than did those born to non-anemic
women (59.9%, SEM = 0.7 versus 55.8%, SEM 1.1, respectively (P = 0.003). One
anemic mother had a newborn with polycythemia (hematocrit > 65%). There was no
difference in SaO2 or respiratory rate in awake or sleeping children, born to mothers with
or without anemia, initially or over the four-month observation period.


View larger version (15K):
[in this window]
[in a new window]
FIGURE 1. Scatter plot and correlation line of mother-newborn pair hematocrits at
3,750 meters.
The mean SaO2 for both awake and asleep states among all the infants decreased through
the first week (P < 0.0001, P = 0.008), then increased significantly (P < 0.0001, P <
0.0001) until four months of age (Figure 2 ). The same trend was found for infants
born to both anemic mothers and those born to non-anemic mothers.

View larger version (15K):
[in this window]
[in a new window]
FIGURE 2. Mean arterial oxygen saturation (SaO2) in infants at 3,750 meters while
awake, feeding, and asleep. Left panel: 10 minutes to 24 hours; right panel: 10 minutes to
4 months. The data of infants born to both anemic women and non-anemic women were
pooled since there was no difference in SaO2 between the two groups.
The SaO2 levels for the infants while awake were significantly greater than for infants
while feeding (P < 0.02) and while asleep (P < 0.01) at all ages studied. Respiratory
frequency in awake infants decreased rapidly after delivery until the 24th hour of life,
reaching a nadir at one week of age (P < 0.0001) and then remaining stable from two
weeks to four months of age (Figure 3 ).

View larger version (14K):
[in this window]
[in a new window]
FIGURE 3. Mean respiratory rate in infants at 3,750 meters while awake and asleep.
Left panel: 10 minutes to 24 hours; right panel: 10 minutes to 4 months. The data of
infants born to both anemic women and non-anemic women are pooled since there was
no difference in respiratory rate between the two groups.

 DISCUSSION
This study demonstrated a high prevalence of anemia in women giving birth to children
of normal birth weight at high altitudes at a Peruvian Andean hospital. In these Peruvian
women living at high altitudes, maternal anemia was not a risk factor for newborn anemia
or hypoxemia.
Maternal anemia is common in both developed and developing countries, yet its effect on
neonatal hematocrit levels has not been well explored. An article from Nigeria18 and
one from Turkey19 demonstrated that low-altitude mothers that are anemic have
neonates that are not anemic. Furthermore, the Turkish study demonstrated that maternal
anemia present in women living at low altitudes was attended by increased fetal
erythropoietin levels and increased erythropoesis.19 Our study in Peru is the first to
demonstrate that at high altitude maternal anemia does not adversely affect either the
infants hematocrit or oxygen content. Moreover, no infant born to an anemic woman was
anemic, demonstrating the ability of the fetus to adapt to maternal anemia even under
conditions of high altitude.
The SaO2 levels of infants born to anemic and non-anemic mothers were similar. Both
the mother and fetus have several mechanisms for adapting to the oxygen demands of a
fetus. Oxygen delivered to the fetus depends on utero-placental blood flow, maternal Hb,
ventilation, and maternal oxygen saturation.
The inverse relationship between maternal and newborn hematocrit suggests that fetal
hematopoiesis can successfully compensate for maternal anemia. The fetus may be
protected from hypoxemia by increased blood flow and improved fetal oxygen extraction
occurring as a response to maternal anemia at high altitudes. Data demonstrating these
mechanisms are only available from studies in sheep.9,10 One of the factors that
may influence this adaptation may be a greater hemodilution during pregnancy at higher
altitudes than that found during pregnancy at sea level, resulting in an enhanced placental
perfusion.12
On the fetal side, at high altitudes newborns have enhanced erythropoiesis and high levels
of blood Hb F, and as a result have much less severe intrauterine hypoxia than do children
born to women native to low altitudes.4,20 After birth, although the neonate has a
high level of Hb F, its high affinity for oxygen may actually limit the release of oxygen
from red blood cells. In contrast, a postnatal increase in the level of 2,3diphosphoglycerate promotes the release of oxygen from Hb and thus lowers SaO2 for a
given partial pressure of arterial oxygen in the first week of life at sea level.21
Infants in our study had SaO2 curves very similar to infants born to both Tibetan women
and to American women living at a high altitude in Leadville, Colorado.6,13 The
decrease in SaO2 during the first week of life and the increase thereafter was consistent
with previous observations with similar experimental designs at 3,658 meters and 3,100
meters.6,13 Our study also demonstrated that throughout the four-months of the
study period, infants born to anemic mothers did not show any difference in their SaO2
levels compared with those born to non-anemic mothers. This would be expected due to
their higher Hb levels.

Infants had higher oxygenation levels while awake compared with when feeding or
sleeping. This was also reported in previous studies in Colorado, Tibet, Bolivia, and Peru.
6,13,22,23 As such, studies on infants at high altitudes need to standardize
values based on the activity of the baby.14
Maternal anemia at high altitudes does not adversely affect hematocrit or newborn
oxygen content among normal weight infants. There is a need for studies in neonates that
examine the mechanism for this successful adaptation at high altitudes.
Received July 31, 2003. Accepted for publication December 10, 2003.
Acknowledgments: We thank Drs. Jaime Zegarra, Manuel Gutierrez, Carlton Evans,
Larry Moulton, Oscar Villegas, and Caryn Bern for their advice, J. B. Phu and D. Sara for
their technical assistance, and the staff of EsSalud Hospital of La Oroya, especially Dr.
Godofredo Pebe Salazar, for their cooperation.
Financial support: This study was supported by the Fogarty-National Institutes of Health
training program International Training and Research in Emerging Infectious Diseases
grant 3D43 TW00910-05S1, National Institute of Allergy and Infectious Diseases TG-35
grant AI-07646-0, and the anonymous RG-ER fund for tropical medicine research.
Authors addresses: Maria E. Ramirez-Cardich, Luis E. Escate, Christopher Kabrhel,
Colleen Johnson, Rivka Galchen, and Christian T. Bautista, A. B. PRISMA, Carlos
Gonzales 251, Urb. Maranga, San Miguel, Lima 32, Peru. Mayuko Saito, St. Lukes Life
Science Institute, 9-1 Akashicho, Chuoku, Tokyo, 104-8560, Japan. Robert H. Gilman,
Department of International Health, The Johns Hopkins School of Pubic Health, 615
North Wolfe Street, Room W3503, Baltimore, MD 21205, Telephone: 410-614-3959,
Fax: 410-614-6060, E-mails: rgilman@jhsph.edu and rgilman@prisma.org.pe .
John J. Strouse, Department of Pediatrics, Division of Hematology, The Johns Hopkins
University School of Medicine, 35 Rutland Road, Baltimore, MD 21205.
REFERENCES
Gamboa R, Marticorena E, 1971. Presin arterial pulmonar en el recin nacido en las
grandes alturas. Arch Inst Biol Andina 4: 5556.[Medline]
Khalid ME, Ali ME, Ali KZ, 1997. Full-term birth weight and placental morphology at
high and low altitude. Int J Gynaecol Obstet 57: 259265.[Medline]
Jensen GM, Moore LG, 1997. The effect of high altitude and other risk factors on birth
weight: independent or interactive effect? Am J Public Health 87: 10031007.
[Abstract/Free Full Text]
Ballew C, Haas JD, 1986. Hematologic evidence of fetal hypoxia among newborn infants
at high altitude in Bolivia. Am J Obstet Gynecol 155: 6669.[Medline]
Zamudio S, Droma T, Norkyel KY, 1993. Protection from intra-uterine growth retardation
in Tibetans at high altitude. Am J Phys Anthropol 91: 215224.[ISI][Medline]
Niermeyer S, Yang P, Shanmina, Drolkar, Zhuang J, Moore LG, 1995. Arterial oxygen
saturation in Tibetan and Han infants born in Lhasa, Tibet. N Engl J Med 333: 1248
1252.[Abstract/Free Full Text]
Moore LG, Zamudio S, Zhuang J, Sun S, Droma T, 2001. Oxygen transport in Tibetan
woman during pregnancy at 3,658 M. Am J Phys Anthropol 144: 4253.

Moore LG, 2001. Human genetic adaptation to high altitude. High Alt Med Biol 2: 257
279.[Medline]
Delpapa EH, Edelstone DI, Milley JR, Balsan M, 1992. Effects of chronic maternal
anemia on systemic and uteroplacental oxygenation in near-term pregnant sheep. Am J
Obstet Gynecol 166: 10071012.[Medline]
Paulone ME, Edelstone DI, Shedd A, 1987. Effects of maternal anemia on uteroplacental
and fetal oxidative metabolism in sheep. Am J Obstet Gynecol 156: 230236.[Medline]

Capurro H, Korichzky S, Fonseca O, Caldeiro-Barcia R, 1978. A simplified method for

diagnosis of gestational age in the newborn infant. J Pediatr 93: 120122.[ISI]
[Medline]
Cohen JH, Haas JD, 1999. Hemoglobin correction factors for estimating the prevalence
of iron deficiency anemia in pregnant women residing at high altitude in Bolivia. Pan Am
J Public Health 6: 392399.
Niermeyer S, Shaffer EM, Thilo E, Corbin C, Moore LG, 1993. Arterial oxygenation and
pulmonary arterial pressure

Supplement
Anemia, Iron and Pregnancy Outcome1 ,2
Theresa O. Scholl3 and Thomas Reilly*
Department of Obstetrics and Gynecology, University of Medicine and Dentistry of New
Jersey-SOM and * Department of Primary Care, University of Medicine and Dentistry of
New Jersey-SHRP, Stratford, NJ 08084
3To whom correspondence should be addressed.
ABSTRACT
When maternal anemia is diagnosed before midpregnancy, it has been associated with an
increased risk of preterm delivery. Maternal anemia detected during the later stages of
pregnancy, especially the third trimester, often reflects the expected (and necessary)
expansion of maternal plasma volume. Third-trimester anemia usually is not associated
with increased risk of preterm delivery. High hemoglobin concentration, elevated
hematocrit and increased levels of serum ferritin late in pregnancy, however, all have
been associated with increased preterm delivery. This increased risk may reflect in part
the failure to expand maternal plasma volume adequately, thus diminishing appropriate
placental perfusion. Although controlled trials of iron supplementation during pregnancy
have consistently demonstrated positive effects on maternal iron status at delivery, they
have not demonstrated reductions in factors that are associated with maternal anemia, i.e.,
increased risk of preterm delivery and infant low birth weight. One reason for discordant
findings may be the exclusion of many gravidas with iron deficiency from these trials or
the data concerning gravidas with pregnancy outcomes such as preterm delivery from the
analysis. Finally, recent concerns have been voiced about harmful effects of iron
supplementation during pregnancy. No adverse effects of iron supplementation on
pregnancy outcome have been demonstrated to date. Questions about the efficacy of iron
supplementation during pregnancy for reducing adverse outcomes such as preterm
delivery and side effects from iron supplementation, including the potential for oxidation
of lipids and DNA, require further research in iron-deficient women.
KEY WORDS: iron anemia hemoglobin ferritin pregnancy outcome

 Maternal anemia during early gestation and poor pregnancy outcome

The relationship between anemia or iron deficiency anemia and increased risk of preterm
delivery (<37 wk gestation) has been supported by several studies (Klebanoff et al. 1991
; Lu et al. 1991 , Murphy et al. 1986 , Scholl et al. 1992 , Scholl and
Hediger 1994 , Zhou et al. 1998 ). Studies have attempted to distinguish actual
iron deficiency anemia from the normal influences of pregnancy-associated hemodilution
as gestation proceeds by studying pregnant women early in gestation.
In a study of 44,000 pregnancies from Cardiff, Wales, Murphy et al. (1986)
examined the prevalence of anemia in women who sought antenatal care by 24 wk
gestation. Increased risk of preterm delivery was associated with low hemoglobin when
women received antenatal care during the first trimester or after midpregnancy (wk 20).
For women beginning antenatal care during the first trimester and between 20 and 24 wk
gestation, an increased risk for preterm delivery was also associated with high
hemoglobin concentrations (>145 g/L). Thus a U-shaped distribution existed for this
cohort, with higher rates of preterm delivery at both ends of the hemoglobin range,
irrespective of the gestation at booking. Although the study population had similar
smoking rates, this factor and others that influence risk for preterm birth were
uncontrolled.
Klebanoff et al. (1991) used a case-control design to examine the relationship
between second and third trimester hematocrit and risk of preterm birth in >1700
gravidas from the Kaiser Permanente Birth Defects Study. For biweekly intervals
between 13 and 26 wk gestation, odds for preterm delivery with anemia were almost
doubled (adjusted odds ratio = 1.9), when controlling for age, education, ethnicity,
marital status, smoking and gestational stage at study entry and consistent for all ethnic
groups. However, during the third trimester, anemia was no longer a risk factor for
preterm delivery.
The relationship between maternal hematocrit and preterm delivery in >17,000 women
receiving iron and folate supplementation in Birmingham, Alabama was reported by Lu
et al. (1991). Before midpregnancy, hematocrit (<37%) was weakly associated with
an increased risk of preterm delivery. This finding, however, was not supported by
multivariate analysis controlling for other risk factors. Hematocrits 40% before 20 wk,
and between 31 and 34 wk gestation were significantly associated with an increased risk
for preterm delivery. Preterm delivery was significantly associated with hematocrits
43% at 3134 wk gestation (odds ratio > 2).
The above studies evaluated maternal hemoglobin concentrations and hematocrits as the
sole indicators assessing iron status during pregnancy. In an effort to distinguish between
iron deficiency anemia and anemias from other causes (e.g., inflammation, infection or
hemodilution) and the risk for preterm delivery, Scholl et al. (1992) reported on 755
pregnant women receiving initial antenatal care at 16.7 5.4 wk gestation in Camden,
New Jersey. These investigators utilized serum ferritin concentrations (<12.0 g/L) to
characterize iron deficiency anemia because anemias from other causes are not associated
with low ferritin concentrations (Institute of Medicine 1990 ). After controlling for
confounding variables, women with iron deficiency anemia early in gestation had more

than a twofold risk for preterm delivery (adjusted odds ratio = 2.66), whereas anemias
from other causes were not associated with any increased risk for preterm delivery.
Previous or current vaginal bleeding at the time of the first antenatal care visit was
documented in 18% of women with anemia. When vaginal bleeding occurred, risk of
preterm delivery was increased fivefold when iron deficiency was present, and more than
twofold when the anemia was the result of other causes. This finding suggested the
possibility that fetal or maternal pathologies influenced the increased risk of preterm
delivery and the vaginal bleeding, which contributed subsequently to the anemia.
In a follow-up study of this population at 28 wk gestation, Scholl and Hediger (1994)
demonstrated that the risk was no longer increased for women who had iron deficiency
anemia (15.6%) at this time or anemia from other causes. Although risk for preterm
delivery was increased when iron deficiency anemia occurs early in gestation, iron
deficiency later in pregnancy probably reflects mainly normal physiologicl expansion of
maternal plasma volume.
These findings were supported by Zhou et al. (1998) who described the relationship
of maternal hemoglobin concentrations during the first trimester and poor pregnancy
outcome in 829 Shanghai women. In this population, other risk factors associated with
poor pregnancy outcome (e.g., smoking or drinking) were uncommon and women
enrolled for care early in the first trimester. Preterm delivery was associated with early
pregnancy hemoglobin concentrations in a U-shaped relationship. The risk of preterm
delivery was increased 1.6 times for women with hemoglobin concentrations between
100 and 109 g/L. A 2.6-fold increase in risk was noted for hemoglobin concentrations
ranging from 90 to 99 g/L. The risk for preterm delivery increased 3.7-fold for
hemoglobin concentrations between 60 and 89 g/L. When hemoglobin concentrations
during mo 5 or 8 of gestation were considered, the risk for preterm delivery was greatly
reduced.
Maternal anemia during gestation and poor pregnancy outcome
Associations between low maternal hemoglobin concentrations and hematocrits at
delivery and poor pregnancy outcome have been reported in several studies. Garn et al.
(1981) reported the effects of maternal anemia on pregnancy in >50,000
pregnancies. Although statistically significant, the relative risk for preterm delivery with
low hematocrit (<29% at any gestation) was modest ( 1.5), except above 25%, where
risk of preterm delivery was doubled for white, but not African-American women. These
data, which were among the first to shed light on anemia and adverse pregnancy
outcome, are compromised by failure to control for confounding variables (stage of
gestation and hemodilution) known to influence the interpretation of hematologic
measurements during gestation.
Using data from >35,000 pregnancies followed in the Collaborative Perinatal Project
(CPP), Klebanoff et al. (1989) concluded that the relationship between maternal
anemia at the time of delivery and preterm delivery was an artifact of blood sample
collection time. During pregnancy, the normal physiologic changes in plasma volume and
red cell mass occur at different periods during gestation. Because these changes are
asynchronous, lower hematocrits typify earlier stages of pregnancy when preterm
delivery commonly occurs, and higher hematocrit values are associated with pregnancies

delivered at later gestational periods. This report did demonstrate a weak association
between anemia early in the third trimester and preterm delivery. After 30 wk, anemia
was not associated with an increased risk of preterm delivery.
High hemoglobin, ferritin and poor pregnancy outcome
Garn et al. (1981) reported hematologic data from the CPP, suggesting that fetal
death, preterm delivery and low birth weight all were increased when the lowest values
obtained during pregnancy exceeded 39% for hematocrit or 130 g/L for hemoglobin.
Steer et al. (1995) confirmed the association, showing that failure of hemoglobin to
fall below 105 g/L was associated with up to a sevenfold increase in risk of low birth
weight and fivefold increase in risk of preterm delivery.
Other findings with high maternal hemoglobin have been observed early in pregnancy, as
well as during the third trimester. Murphy et al. (1986) found that at every
gestational stage examined, high hemoglobin (>133 g/L)was associated with increased
risk of one or more poor outcomes. High booking hemoglobin was associated positively
with maternal hypertension, a relationship that was evidenced as early as the first
trimester. In general, most associations tended to be stronger later in pregnancy than in
the first trimester, implicating plasma volume expansion.
Zhou et al. (1998) examined high hemoglobin along with anemia in their
observational study. At entry to care, which ranged between 6 and 8.4 wk gestation,
women with hemoglobin levels exceeding 130g/L had a greater than twofold increase in
risk of preterm delivery and infant low birth weight. Neither risk was statistically
significant, however, because of the small numbers with high hemoglobin.
Similarly, a concentration of the iron storage protein, ferritin, that is high for the third
trimester of pregnancy is also associated with an increased risk for preterm and very
preterm delivery. From their studies of Alabama women, both Tamura et al. (1996)
and Goldenberg et al. (1996) found high third-trimester ferritin levels (>40 ng/L) to
be a marker for an increased risk for preterm and very preterm delivery. Prospective data
from Camden (Scholl 1998 ) indicated that high ferritin levels (>41.5 ng/L) during
the third trimester, stemming from the failure of ferritin to decline from entry, increased
risk of very preterm delivery more than eightfold. High ferritin also was associated with
indicators of infection, including clinical chorioamnionitis (Scholl 1998 ) and infant
sepsis among women with pregnancies complicated by premature rupture of membranes
(Goldenberg et al. 1998 ). A hallmark of maternal hypovolemia, i.e., high maternal
hemoglobin (>120 g/L), was more frequent among women with high ferritin (Scholl 1999
) during the latter half of pregnancy. Thus, like hemoglobin, failure of the plasma
volume to expand or hypovolemia also is implicated in the etiology of high maternal
ferritin.
It is possible that anemia or other factors related to maternal nutritional status early in
pregnancy are associated with later hypovolemia. Poorly nourished animals have reduced
maternal plasma volume expansion during pregnancy and low cardiac output, with lower
uteroplacental blood flow and nutrient transmission to the fetus (Rosso and Salas 1994
). Similarly, Camden women with high third trimester ferritin had numerous indicators
of poor nutritional status earlier in pregnancy [risk of anemia and iron deficiency anemia
were increased, circulating levels of ferritin and serum and red cell folate were lower

(Scholl 1998 )]. Later on (wk 28), their profiles suggested hypovolemia, i.e., high
hemoglobin was more frequent and anemia and iron deficiency anemia were less
common than in controls (Scholl 1999 ).
Iron supplementation and pregnancy outcome
Controlled trials of iron supplementation during pregnancy have consistently
demonstrated positive effects on maternal iron status at delivery. The prevalence of low
hemoglobin or hematocrit is reduced; serum ferritin, serum iron and almost every other
measure of maternal iron status, including bone marrow iron, are increased in comparison
with controls (Mohamed 1998 ). However, the same benefits are not readily
demonstrable when iron supplementation is incorporated into routine prenatal care. The
estimated prevalence of third-trimester anemia that would occur in the absence of iron
supplementation is estimated at 40%, compared with the 37% reported by the Centers for
Disease Control (CDC) surveillance of low income women (Perry et al. 1995 , Yip
1996 ).
Controlled trials of iron supplementation have not demonstrated reductions in factors that
are associated with maternal anemia, i.e., increased risk of preterm delivery and infant
low birth weight. However, effects on pregnancy outcome have been difficult to evaluate
because few studies have addressed the effect of iron supplementation in groups in which
anemia and iron deficiency anemia are prevalent (Mohamed 1998 ). In one metaanalysis examining differences in infant birth weight and morbidity in trials from Western
Europe, five of six trials excluded women with low hemoglobin (<100g/L) at entry,
women already taking iron supplements (who were likely to be iron deficient) and those
with a prior poor pregnancy outcome. In three of six trials, patients with characteristics
related to the outcome of interest (e.g., gravidas developing anemia or delivering
preterm) were excluded (Hemminiki and Starfield 1978 ).
The trial of routine vs. selective iron supplementation (Hemminiki and Rimpela 1991a
and 1991b ) also focused on nonanemic women (hemoglobin >100 g/L at entry).
Routine supplementation resulted in a reduced risk of infant low birth weight (odds ratio
= 0.89) and preterm delivery (odds ratio = 0.71), which were not significant (Hemminiki
and Rimpela 1991a and 1991b , Mohamed 1998 , Mohamed and Hytten
1989 ). Thus, although the conclusion that iron supplementation has no effect on the
outcome of pregnancy may be true, its efficacy has been evaluated almost exclusively
among women who were not anemic early in pregnancy and were therefore less likely to
realize the potential benefits of supplementation.
Some trials conducted among women from the developing world, in which anemia and
presumably iron deficiency are prevalent, have come to somewhat different conclusions
albeit from smaller numbers and often with substantial loss to follow-up and potential
bias to the results of the trial. Preziosi et al. (1997) , for example, supplemented
Nigerian women with elemental iron (100 mg) during the third trimester and found
improvements in indices of maternal iron status, birth length and Apgar scores but no
difference in infant birth weight. Agarwal et al. (1991) randomized women from six
subcenters in an Indian district to iron/folate (60 mg elemental iron + 500 g folic acid)
or placebo. After excluding preterm deliveries, they reported improvements in infant birth
weight and term low birth weight with supplementation.

 Adverse effects of iron supplementation

One concern voiced about iron supplementation during pregnancy is that because iron
allays the fall in hemoglobin during pregnancy, iron-induced macrocytosis could increase
blood viscosity to a degree that would impair uteroplacental blood flow, decrease
placental perfusion and increase risk of placental infarction (Koller 1982 ,
Mohamed and Hytten 1989 ).
Hemminiki and Rimpela (1991a and 1991b) conducted a clinical trial of
selective vs. routine iron supplementation in 2912 Finnish women. Data from that
multicenter study were examined to determine whether routine supplementation with iron
in nonanemic women increased risk of high maternal hemoglobin and poor fetal growth.
Women randomized to the selective iron group received iron supplements only when
hematocrit fell below 30% or hemoglobin below 100 g/L on two consecutive visits after
week 33.
Routine supplementation with iron did increase maternal hematocrit. In selectively
supplemented women, hematocrit declined from wk 12 to 28, whereas in routinely
supplemented women, the decline was arrested by wk 20. Although routine
supplementation increased hematocrit, it did not alter infant birth weight. In contrast,
gestation duration was increased significantly (+ 0.2 wk). Interestingly, in both routine
and selectively supplemented groups, hematocrit was negatively correlated with birth
weight and placental weight. This correlation was first evidenced at baseline (i.e., wk 12
gestation) and persisted after adjusting for the effect of maternal blood pressure. Thus,
rather than iron supplementation, factors that are intrinsic to pregnancy (poor plasma
volume expansion, increased blood viscosity) and that are evidenced early in gestation
seem to link high hemoglobin with poor pregnancy outcome.
Scholl (1999) and Scholl and Schroeder (1999) examined the influence of
elemental iron supplement use on high third-trimester ferritin levels and risk of preterm
and very preterm delivery. Both anemic and nonanemic gravidas who used elemental iron
supplements by wk 28 were compared with gravidas without such supplement use.
Anemia was assessed at entry to care (15 4.9 wk) by hemoglobin concentration using
the CDC criteria for pregnancy (Centers for Disease Control and Prevention 1989 ).
After control for potential confounding variables and in comparison to gravidas who did
not use iron, anemic women had a significantly (>threefold) increased risk of high ferritin
at wk 28 when they used iron. Nonanemic women who used iron also sustained a
significant increase in the odds of a high ferritin level; in this case, risk was increased
twofold in comparison to controls (Table 1 ). Interestingly, the absolute risk of
having a high third-trimester ferritin level was greater among the anemic than among
nonanemic women using iron (Table 1) . Consistent with the findings of Hemminiki
and Rimpela (1991a and 1991b) on iron supplementation and birth weight,
elemental iron use did not increase risk of preterm or very preterm delivery in either
anemic or nonanemic gravidas (Table 1) .

Another current concern is that iron supplements are a possible source of free radical

development with the potential to cause oxidative damage to DNA, lipids and protein
(Knutson et al. 1999 , Lund et al. 1999 ). Apart from the association of
preeclampsia with oxidative stress (Walsh 1998 ), this is an area that has been little
explored during pregnancy. Oxidative stress is believed to be involved in the etiology of
cardiovascular disease, cancer, cataracts, inflammatory diseases, immune function and
numerous other disorders (Gutteridge and Halliwell 1994 ). During pregnancy,
oxidative stress from iron supplementation has the potential to damage the conceptus,
increasing risk of congenital defects, preterm delivery and low birth weight in the short
term.
Clearly many issues surrounding iron supplementation during pregnancy must be
addressed, including the appropriate window in which supplements may have maximum
effect on the associated risks and compliance with the use of iron supplements. Assessing
deficits in iron stores also would be important for identifying women who are iron
deficient and most likely to be responsive to iron supplementation. If anemia is in fact
due to iron deficiency and is causally related to preterm delivery, then iron
supplementation in the appropriate window should reduce that risk. Viteri (1997)
suggests providing menstruating women at risk with weekly iron/folate tablets, an
intervention likely to improve a womens iron status before conception and reduce the
risk of iron deficiency anemia in early pregnancy.
Apart from a few trials in the developing world in which loss to follow-up is problematic,
questions about the efficacy and hypothetical side effects of iron supplementation (e.g.,
greater oxidative stress or impaired utero-placental blood flow) have not been addressed
with the use of a randomized, double-blind design in an appropriate population. Such a
population is one in which anemic women are likely to be iron deficient, one in which
iron supplementation is not the norm because of ethical considerations about withholding
treatment, and one in which women can be followed and monitored until they deliver
because of the potential bias and misinterpretation associated with substantial loss to
follow-up.
FOOTNOTES
1 Presented at the symposium entitled "Improving Adolescent Iron Status before
Childbearing" as part of the Experimental Biology 99 meeting held April 1721 in
Washington, DC. This symposium was sponsored by the American Society for
Nutritional Sciences and was supported in part by an educational grant from
Micronutrient Initiative. The proceedings of this symposium are published as a
supplement to The Journal of Nutrition. Guest editors for the symposium publication
were Kathleen Kurz, International Center for Research on Women and Rae Galloway,
World Bank/Micronutrient Initiative.

2 Supported by grants HD18269 and ES 07437 from the National Institutes of Health.

REFERENCES

1. Agarwal K. N., Agarwal D. K., Mishra K. P. Impact of anaemia prophylaxis in

pregnancy on maternal haemoglobin, serum ferritin and birth weight. Indian J. Med. Res.
1991;94:277-280[Medline]
2. Centers for Disease Control and Prevention CDC criteria for anemia in children and
childbearing-aged women. Morb. Mortal. Wkly. Rep. 1989;38:400-404[Medline]
3. Garn S. M., Ridella S. A., Petzoid A. S., Falkner F. Maternal hematologic levels and
pregnancy outcomes. Semin. Perinatol. 1981;5:155-162[Medline]
4. Goldenberg R. L., Mercer B. M., Miodovnik M., Thurnau G. R., Meis P. J., Moawad
A., Paul R. H., Bottoms S. F., Das A., Roberts J. M., McNellis D., Tamura T. Plasma
ferritin, premature rupture of membranes, and pregnancy outcome. Am. J. Obstet.
Gynecol. 1998;179:1599-1604[Medline]
5. Goldenberg R. L., Tamura T., DuBard M., Johnston K. E., Copper R. L., Neggers Y.
Plasma ferritin and pregnancy outcome. Am. J. Obstet. Gynecol. 1996;175:1356-1359
[Medline]
6. Gutteridge J.M.C., Halliwell B. Antioxidants in Nutrition, Health and Disease 1994
Oxford University Press New York, NY.
7. Hemminki E., Rimpela U. Iron supplementation, maternal packed cell volume, and
fetal growth. Arch. Dis. Child. 1991a;66:422-425[Abstract/Free Full Text]
8. Hemminki E., Rimpela U. A randomized comparison of routine versus selective iron
supplementation during pregnancy. J. Am. Coll. Nutr. 1991b;10:3-10[Abstract]
9. Hemminiki E., Starfield B. Routine administration of iron and vitamins during
pregnancy: review of controlled clinical trials. Br. J. Obstet. Gynaecol. 1978;85:404-410
[Medline]
10. Institute of Medicine, Committee on Nutritional Status during Pregnancy and
Lactation Nutrition during Pregnancy 1990:272-298 National Academy Press
Washington, DC.
11. Klebanoff M. A., Shiono P. H., Berendes H. W., Rhoads G. G. Facts and artifacts
about anemia and preterm delivery. J. Am. Med. Assoc. 1989;262:511-515[Abstract]
12. Klebanoff M. A., Shiono P. H., Selby J. V., Trachtenberg A. I., Graubard B. I. Anemia
and spontaneous preterm birth. Am. J. Obstet. Gynecol. 1991;164:59-63[Medline]

Maternal Anemia: A Preventable Killer

Iron deficiency is one of the most

prevalent nutritional deficiencies in
the world and is reported by the
World Health Organization (WHO) to
affect four to five billion people. WHO
estimates that two billion people
suffer from anemia. Approximately
50% of all anemia is estimated to be
due to iron deficiency, a condition
of deteriorating iron reserves in the
body caused by low dietary intake
of iron, poor absorption of dietary
iron, or blood loss (for example, from
hookworm, repeated childbirth or
heavy menstruation) which leads to
loss of iron. Iron deficiency anemia
(IDA) is the most severe form of iron
deficiency, and results when the bodys
iron supply cannot support production
of hemoglobin in adequate amounts
to maintain normal functioning of the
body. Anemia from other causes
(and therefore, not iron deficiency
anemia), results from malaria or from
genetic disorders, among other causes.
Other micronutrient deficiencies (e.g.,
vitamins A, B6 and B12, riboflavin, and
folic acid) are also known to cause
anemia (Figure 1).
Anemia and iron deficiency remain
at epidemic levels among women and
children in many nations. Given the
availability of proven interventions to
prevent and treat anemia caused by a
variety of determinants, the persistent
high prevalence represent a lack of
political will and failure of the public
health sector. New estimates of the
Iron is critical to oxygen transport in the body
In the human body, iron is present in all cells and has several vital functions
as a carrier of oxygen to the tissues from the lungs in the form of hemoglobin
(Hb); as a facilitator of oxygen use and storage in the muscles as myoglobin
and as an integral part of enzymes. Anemia is defined as a low level of Hb in
blood.
numbers of maternal and perinatal
deaths associated with iron deficiency

anemia underscore the urgent need

to refocus resources and public health
priorities to more effectively tackle the
problem.
1. Anemia and iron deficiency
are highly prevalent conditions
with major consequences for
health, survival, and economic
development
Anemia prevalence is highest among
pregnant women, infants, and young
children due to the high iron demands
of growth and pregnancy. On average,
45% of pregnant women and 49% of
1 Mason, Rivers and Helwig. Recent trends in malnutrition in developing regions:
Vitamin A deficiencies,
anemia, iodine deficiency, and child underweight, Food and Nutrition Bulletin 26: 57162, 2005.
chidren under five years of age are
anemic in developing regions.1 The
consequences of anemia include:
Increased maternal and perinatal
mortality
Increased numbers of preterm
birth and/or low birthweight
Impaired cognitive development in
children
Reduced adult work productivity
Hence, anemia prevention programs
can contribute significantly to achieving
many of the Millennium Development
Goals (MDGs) including MDGs 1
(Poverty and hunger); 2 (Universal
education); 4 (Child mortality
reduction); and 5 (Improved maternal
health).
2. Iron deficiency is high on
the list of risk factors for
global maternal and perinatal
mortality
WHOs 2002 Global Burden of Disease
Report identified iron deficiency as
the 12th most important risk factor
for all mortality globally, and the 9th
most important risk factor for the
Iron Deficiency

Anemia
Iron Deficiency
Anemia
Figure 1: Iron deficiency, iron deficiency anemia (IDA) and anemia
CAUSES
Low dietary
intake of iron
Poor absorption
of iron
Malaria
Hookworm
High fertility
Other
micronutrient
deficiencies
Diarrhea, HIV/
AIDS and other
infectious diseases
CONSEQUENCES
Increased
maternal and
perinatal
mortality
Increased
numbers of preterm
birth and/or
low birthweight
Impaired
cognitive
development
Reduced work
productivity
global burden of disease. Recent
WHO analysis of causes of maternal
death showed that hemorrhage is the
major contributor to maternal deaths
in developing countries.2
In a separate analysis, iron deficiency
anemia (IDA) was an underlying risk
factor for maternal and perinatal
mortality and morbidity, and was
estimated to be associated with
115,000 of the 510,000 maternal
deaths (22%) and 591,000 of the
2,464,000 perinatal deaths (24%)
occurring annually around the world

(Figure 2). 3, 4
The consequences of anemia are
serious:
Anemia in pregnant women
reduces a womans ability to
survive bleeding during and
after childbirth (i.e., post partum
hemorrhage (PPH)) and may
result in premature and/or lower
birthweight babies with a higher
risk of death
Iron deficiency with or
without anemia limits cognitive
development in children. It reduces
their achievement in school and
ultimately undercuts the benefits
of investing in education.
Anemia and iron deficiency cause
weakness, fatigue, and reduced
physical ability to work. Economic
analysis show that for every $1
spent on iron supplementation
programs for pregnant women,
there is a return of US $24 in
decreased disability and increased
wages over a womans lifetime.4
3. New analyses show that mild
and moderate not just severe
anemia has serious consequences
for women and children
A recent meta-analysis3 shows that
correcting anemia of any severity
reduces the risk of death: the risk
of maternal mortality decreases
by about 20% for each 1 g/dL
increase in Hb. This decreased risk
is continuous over the full range of Hb
between 5 and 12 but it is not linear
the decrease in risk is greater at
the lower Hb concentrations. This is
a new finding and different from the
earlier view that only severe anemia
is associated with increased mortality.
It has important policy and program
implications:
More women are affected by

increased mortality risk. Previous

estimates of the number of women at
increased risk of death associated with
anemia only considered those women
with severe anemia. In light of the
meta-analysis results, more pregnant
women are estimated to be at risk,
since the majority of anemic women
have hemoglobin concentrations
between 7 and 12 g/dL (mild and
moderate anemia) compared to the
relative few with severe (Hb < 7 g/dL)
or very severe anemia (Hb < 5 g/dL).
Hence programs should focus on mild
and moderate anemia, in addition to
severe anemia, for public health
impact.
Anemia/low hemoglobin
creates cardiovascular
problems
Underpinning the relationship
between hemoglobin
concentration and mortality
risk is the fact that death
from cardiovascular causes is
a function of blood volume,
blood loss, cardiac fitness, and
hemoglobin concentration.
2 Calculated from WHO Analysis of Causes of Maternal Deaths: A Systematic
Review, The Lancet 367:
1066-1074, 2006.
3 Stoltzfus, Mullany and Black. Iron Deficiency Anemia, Comparative quantification of
health risks: Global and
regional burden of disease attributable to selected major risk factors, WHO 2004.
4 Based on the quantitative relationship, one can estimate the percent of maternal
mortality attributable to
iron deficiency anemia (IDA) in a given country when maternal mortality ratio,
prevalence of anemia, and
number of births per year are known (PROFILES, IDA Calculator at
www.fantaproject.org).
2
Hemorrhage
31%
HIV 3%
Indirect
Causes

14%
Unsafe Abortion
5%
Unclassified
6%
Obstructed
Labor 7%
Other Direct
Causes 5%
Sepsis 11%
Hypertensive
Disorder 10%
Anemia
8%
Deaths
Associated
with IDA
22%
Figure 2. Causes of Maternal Death2 and Contribution of Iron
Deficiency Anemia (IDA)3
6 Yip. Iron supplementation during pregnancy: Is it effective? American Journal of
Clinical Nutrition 63: 835-855, 1996.
7 Mannar and Gallego. Iron fortification: Country level experiences and lessons
learned, Journal of Nutrition 132:856S-858S, 2002.
8 Strategic framework for malaria control during pregnancy in the WHO Africa region.
WHO 2003.
9 Report of the WHO Informal Consultation on Hookworm Infection and Anemia in
Girls and Women, WHO, 1996.
10 Conde-Agudelo and Belizan. Maternal morbidity and mortality associated with
interpregnancy interval: Cross sectional study, British Medical Journal 321: 1255-1259,
2000.
11 Soekarjo, de Pee, Kusin and Bloem. School-based supplementation: Lessons learned
in Indonesia, Standing Committee on Nutrition News 31, 2005.
12 Abel, Rajaratnam, Kalaimani and Kirubakaran. Can iron status be improved in each
of the three trimesters? A community-based study. European Journal of Clinical
Nutrition 54: 490-493, 2002.
Iron Deficiency Malaria Hookworm High Fertility HIV/AIDS
Sub-Saharan Africa l l v l v
South and SouthEast Asia l v t v t
North Africa l t t v t
Americas l t t t t
Central Asia/Caucasus l v t v t
Western Pacific (includes China) l t v t t
Adapted from: Galloway, R. Anemia Prevention and Control: What Works. Washington,
DC: USAID, June 2003.
Table 1: Causes of Anemia - Relative Importance by Region Relative

Importance
by Region
Key:
l high
v medium
t low
4. Knowing the various causes of
anemia in a target population is
the first step in designing tailored
intervention strategies
The main causes of anemia include:
inadequate intake and poor
absorption of iron
malaria, particularly in young
children and pregnant women
hookworms
diarrhea, HIV/AIDS and other
infections
genetic disorders (e.g., sickle cell
and thalassemia)
blood loss during labor and
delivery; heavy menstrual blood
flow; closely spaced pregnancies
Table 1 below highlights the relative
importance of these causes by region.
The relative importance of various
causes differs by age and sex of the
population and setting. The greatest
burden of death and disease due to
anemia is in Africa and Asia and is
associated with the consequences of
anemia among pregnant women and
young children.
5. Proven interventions are
available to address the major
causes of anemia in women
There is consensus on a number of
evidence-based interventions to reduce
anemia prevalence among pregnant
women/women of reproductive age
when they are applied effectively to
a population with known causes of
anemia.
Universal supplementation of
pregnant women with daily iron
folic acid tablets;6

 Fortification of commonly
consumed food products with
micronutrients;7
Control of malaria by intermittent
preventive treatment (IPT), longlasting
insecticide treated bed nets
(ITN), indoor residual spraying
(IRS), and Artemisinin-Containing
Antimalarial Combination Therapy
(ACT);8
Control of hookworms by
deworming medication such as
albendazole and mebendazole
as a routine part of ANC where
hookworm prevalence >20%;9
Optimal birth spacing;10 and
Programs that improve the iron
stores of adolescents with weekly
iron/folic acid supplements.11
6. Maternal anemia programs
are most effective when they
address the multiple causes
of anemia through integrated
interventions12
Programs need to address the primary
preventable causes of anemia in a
coordinated way. The global Roll
Back Malaria initiative has worked in
tandem with maternal health and safe
motherhood programs since 1998 to
prevent and control anemia caused by
malaria in pregnant women through
integrated program interventions.
Efforts to address maternal anemia
should be tailored to the local context
and situation and take advantage
of partnerships with agencies and
programs that are implementing
relevant interventions. Presently,
there are several global initiatives that
focus on prevention of anemia:
The Global Alliance for Improved
Nutrition (GAIN a fortification
initiative at www.gainhealth.org)
The Presidential Malaria Initiatives
(PMI, www.fightmalaria.gov)

 Partners for Parasite Control

(www.who.wormcontrol/en/)
Global Fund for AIDS, Tuberculosis
and Malaria (www.globalfund.org)
Presidents Emergency Plan for
AIDS Relief (www.state.gov/s/gac)
3
This report is the product of a collaboration among USAIDs A2Z Micronutrient and
Child Blindness
Project, ACCESS Program, and Food and Nutrition Technical Assistance (FANTA)
Project. This report
is made possible by the generous support of the American people through the Office of
Health,
Infectious Disease, and Nutrition, Bureau for Global Health, United States Agency for
International
Development (USAID). The A2Z and FANTA Projects are implemented by the Academy
for
Educational Development (AED). The ACCESS Program is implemented by JHPIEGO.
The contents do
not necessarily reflect the views of USAID or the United States Government.
FOOD AND
NUTRITION
TECHNICAL
ASSISTANCE
USAID health and nutrition programs
that address maternal anemia include:
A2Z Micronutrient and Child
Blindness Project,
Access to Clinical and Community
Maternal, Neonatal, and Womens
Health Services (ACCESS)
Program, and
Food and Nutrition Technical
Assistance (FANTA) Project.
7. Success with national level
anemia prevention programs has
been documented
In Nicaragua the prevalence of anemia
in reproductive-age women steadily
decreased from 1993 to 2003, going
from 34% to 16% (Figure 3). Hallmarks
of the Nicaragua anemia prevention
program included the following
Iron was distributed through
antenatal care (ANC) clinics and
community volunteers did followup

and counseling work to support

women.
Supplies of iron/folic acid
supplements (IFA) were adequate
and a strong behavior change
communication (BCC) program
was implemented.
87% of women reported taking
IFA in their last pregnancy, and of
those who took IFA, 53% took it
for longer than 6 months.
Side-effects were not reported
as a major issue. Women were
counseled to expect side-effects
and reported knowing how to
cope with them.
Universal wheat flour fortification
was introduced in 1997; the
introduction of vitamin A sugar
fortification may have also
contributed to anemia reduction.
India recently intensified its anemia
control programs for adolescent girls,
ages 10 to 19 years, in seven states.
Diets with low bioavailable iron
combined with helminth infections
are the main causes of anemia among
girls and women in India.13 After one
year of program implementation,
anemia prevalence decreased in all
states, varying from reductions of 5
percentage points in Jharkhand to 50
percentage points in Uttar Pradesh.
In Andra Pradesh, anemia prevalence
decreased 43 percentage points at one
year and 70 percentage points after two
years of the program implementation.
The components of the anemia
programs included:
Weekly IFA
Abendazole to treat worm
infestation
Behavior change communication
that covered the importance of
nutrition including the intake of
iron/folic acid tablets

8. Progress of anemia programs

is being monitored
To monitor the progress of anemia
programs at the national level, the
Demographic and Health Survey
(DHS) has been collecting data on
anemia prevalence and coverage of
iron supplementation during pregnancy
since the mid-1990s. DHS also collects
data on malaria programs, such as ITN
use, use of IPT by pregnant women and
starting in 2006, DHS will also obtain
information on the use of drugs for
intestinal worms. Combined data on
these indicators will assist countries to
assess and improve anemia prevention
and control programs.
Information on the following key
indicators is available for selected
countries from DHS:
Anemia prevalence among children
and women of reproductive age
Iron supplementation during
pregnancy
Use of ITN by children and
pregnant women
Use of IPT by pregnant women
Use of deworming drugs among
children and pregnant women
Moving forward with a package of
proven interventions to prevent
and control maternal anemia
It is clear from experiences that
countries implementing a package
of proven interventions are much
more likely to succeed in improving
maternal anemia than countries
implementing any one of the single
interventions. Hence countries are
encouraged to plan for an integrated
package of interventions that may
include iron supplementation of
pregnant women and adolescent girls,
fortification of food with iron, malaria
and hookworm control and optimal
birth spacing as appropriate.

13 Dwidedi and Schultink. Reducing anemia

among Indian adolescent girls through once-weekly
supplementation with iron and folic acid, SCN News
31:19-23, 2006.
August 2006
4
FIGURE 3. Nicaragua: Prevalence
of anemia in mothers/caregivers
1993, 2000 & 2003
Source: SIVIN 2004

33.6
23.7
16.1

Percentage Hb < 12 g/dL

Review article

Progestogen-only contraceptive use among

women
with sickle cell anemia: a systematic review
Jennifer K. LegardyT, Kathryn M. Curtis
WHO Collaborating Center in Reproductive Health, Division of Reproductive Health, Centers for Disease
Control and Prevention, Atlanta, GA 30341, USA
Received 27 July 2005; accepted 11 August 2005

Abstract
The use of progestogen-only contraceptives among women with sickle cell anemia has
generated concerns about possible hematological
and other clinical complications. Based on the literature, we assessed whether use of
progestogen-only contraceptives is associated with
adverse health effects among women with sickle cell anemia. We searched the MEDLINE
database for articles published in peer-reviewed
journals between 1966 and September 2004 that were relevant to sickle cell anemia and use
of progestogen-only contraceptives. Of the
70 articles identified through the search, 8 met the criteria for this review. These studies did
not identify any adverse events or clinically or
statistically significant adverse changes in hematological or biochemical parameters
associated with the use of progestogen-only
contraceptive methods. Six studies suggested that users experienced a decrease in clinical
symptoms and less frequent and severe painful
crises compared with nonusers. Although data are limited, these studies suggest that
progestogen-only contraceptives are safe for women
with sickle cell anemia.
D 2006 Elsevier Inc. All rights reserved.
Keywords: Sickle cell anemia; Progestogen; Contraception; Systematic review

1. Introduction
Sickle cell disease is caused by the pairing of an inherited
autosomal recessive gene (h-globin), which affects the red
blood cells [1,2]. Deoxygenation of the red blood cells
causes these cells to change from their normal round shape
to a rodlike sickle shape. These sickle-shape cells adhere to
the blood vessels, eventually clogging the vessels and
blocking normal flow of blood and oxygen to organs and
tissue. Of the several forms of this disease, the most
common and severe is sickle cell anemia. Primarily, people
of African and Mediterranean ancestry are affected by the
disease and often are carriers of the trait. Based on statistics
from the National Heart, Lung, and Blood Institute, in the
United States, approximately 1 in every 500 African
American babies and 10001400 Hispanic American babies
are diagnosed with this disease each year. The sickle cell
trait (one copy of the gene that causes sickle cell anemia) is
present in an estimated 2 million Americans [3]. The
medical complications that arise from this disease include
chronic anemia, episodes of musculoskeletal pain or
bcrises,Q acute chest syndrome and stroke. In addition,
sickle cell disease in women has been associated with
significant maternal morbidity and mortality [4,5].
Using contraception and becoming pregnant are major
decisions for a woman with sickle cell anemia because of
the related medical complications. Historically, it has been

recommended that women with sickle cell anemia avoid

pregnancy, become sterile or have an abortion because of
the increased risk of maternal and fetal mortality [4,5].
Maternal and perinatal mortality rates associated with sickle
cell disease have declined significantly over time, due to
advancements in medical technology and obstetric and
perinatal care. However, a recent study showed that women
with sickle cell anemia still have an increased risk of
pregnancy complications, such as premature rupture of the
membranes, preterm labor, antepartum admission and
postpartum infections [5].
Women with sickle cell anemia who want to avoid
pregnancy need to have appropriate and reliable counseling
regarding contraceptive use. However, there have not been
any clear universal guidelines established for contraceptive
0010-7824/$ see front matter D 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.contraception.2005.08.010
T Corresponding author. Tel.: +1 770 488 6513; fax: +1 770 488
6391.
E-mail address: jlegardy@cdc.gov (J.K. Legardy).
Contraception 73 (2006) 195 204

use among these women. Traditionally, sickle cell disease

has been considered a contraindication to the use of
combined oral contraceptives (COCs), especially in Europe
[6]. Yet, evidence is limited regarding the effects of COC
use in this population [2,68]. The World Health Organization
(WHO) currently classifies sickle cell anemia as a
Category 2 for COCs, meaning that the benefits of COC use
among women with this condition generally outweigh the
theoretical or proven risks [9]. There have also been
concerns about the safety of using intrauterine devices
(IUDs) for women with sickle cell disease, but studies have
shown that copper IUDs are a safe method of contraception
for these women [6,10]. In addition, WHO classifies sickle
cell anemia as Category 1 for IUD use, meaning there is no
restriction for their use [9]. For progestogen-only contraceptives
(i.e., pills, injectables and implantables), possible
hematological and other clinical complications are the major
concerns regarding the use of these contraceptives among
women with sickle cell anemia.
We conducted this systematic review in preparation for
an Expert Working Group of international family planning
experts convened by WHO in October 2003 to develop and
revise medical eligibility criteria for contraceptive use. In
this report, we evaluate the scientific evidence regarding
whether women with sickle cell anemia experience adverse
health effects while using progestogen-only contraceptives,
as well as provide the WHO recommendations that were
derived in part from this evidence. We considered the
following progestogen-only contraceptive methods: Norplant
and Norplant-2 or Jadelle (levonorgestrel-releasing
contraceptive implants), Implanon (etonogestrel-releasing
contraceptive implants), Uniplant (nomegestrol acetatereleasing
contraceptive implants), depot medroxyprogesterone
acetate (DMPA) injectables, norethisterone enanthate
(NET-EN) injectables and progestogen-only contraceptive

pills (any formulation).

2. Materials and methods
We searched the MEDLINE database for articles (in all
languages) published in peer-reviewed journals from 1966
to September 2004. Search terms included bsickle cell
diseaseQ or banemiaQ along with the following terms: (a)
depot and medroxyprogesterone 17-acetate or medroxyprogesterone
acetate, (b) DMPA, (c) progestogen-only contraception
or progestational hormones or progestogen and only
and contraceptive agents, (d) norethisterone-enanthate, (e)
NET EN, (f) contraception, (g) contraception or female
contraception, (h) implants, (i) Norplant, (j) Uniplant, (k)
Jadelle, (l) Implanon, (m) levonorgestrel, (n) etonogestrel,
and (o) effects and contraceptive agents or contraceptives.
We also searched the reference lists of the articles
identified by the electronic search and letters to the editor
to identify additional relevant articles. We did not consider
unpublished manuscripts, dissertations or abstracts from
scientific conferences.
2.1. Study selection
Seventy articles were identified by searching MEDLINE
and by reviewing reference lists. Eight of the 70 articles
matched the goal of the review, which was to evaluate health
outcomes related to progestogen-only contraceptive use
among women with sickle cell anemia. These articles were
selected after reviewing the titles, abstracts and articles. The
eight articles examined the safety of using different
progestogen-only contraceptive methods among women
with sickle cell anemia, including clinical and other
hematological effects.
2.2. Study quality assessment
We summarized and systematically assessed the evidence
using a standard abstract form. Each article was reviewed
for quality using a preliminary draft of the grading system
developed by the Grades of Recommendation Assessment,
Development and Evaluation (GRADE) Working Group
(Appendix A) [11].
2.3. Data synthesis
This review includes two randomized controlled trials
(RCTs), two non-randomized controlled trials, two prospective
cohort studies and two cross-sectional studies (Table 1).
Due to the overall low prevalence of sickle cell anemia in the
general population, the sample sizes of these studies were
small. The studies reported means and standard deviations,
percentages or qualitative results, so summary odd ratios
were not calculated.
3. Results
3.1. Implantables (Norplant, Norplant-2, Jadelle and
Uniplant)
A prospective cohort study conducted in Nigeria
evaluated the safety of Norplant use among 25 women
with homozygous sickle cell anemia [12]. These women
were followed for a mean of 12 months after insertion.
Outcomes were measured pre- and postinsertion for each
woman, and the women served as their own controls. At

baseline, these women had not experienced a painful crisis

or serious infection in the past month; had no clinical
evidence of gall bladder disease; and had no history or
evidence of aplastic crisis, diabetes, bleeding disorders,
myocardial infarction, stroke or severe liver, renal, skeletal,
pulmonary, skin or ocular disease. Insertion of the Norplant
did not cause any serious or unexpected adverse side
effects. By using paired t tests, this study showed no
statistically significant changes after Norplant was inserted
in the means of any of the five biochemical and six
hematological parameters measured. For example, the fetal
hemoglobin (HbF) percent (hematological parameter) preand
postinsertion mean was 3.1 and 3.4, respectively. For
the serum albumin (biochemical parameter), the pre- and
postinsertion mean was 3.8 and 3.9, respectively. This study
J.K. Legardy, K.M. Curtis / Contraception 196 73 (2006) 195204

had a high retention rate (23/25 participants) that allowed

for 80% power to detect an adverse event occurring in 7%
of the participants.
In two studies conducted in Brazil, which examined the
effects of Uniplant, no adverse effects were reported among
women with sickle cell anemia. Ten otherwise healthy
women of reproductive age with sickle cell anemia were
enrolled in a 1-year, prospective cohort study to examine the
effects of using Uniplant on carbohydrate metabolism [13].
These participants had no contraindications to hormonal
contraceptives and had not used a hormonal contraceptive in
the past 6 months. The outcomes of interest were glucose
tolerance and insulin levels, measured pre- and postinsertion,
and the women served as their own controls. Participants did
not experience any changes in these outcomes during the
study period. The womens fasting glycosylated hemoglobin
levels before insertion ranged from 77 to 90 mg/dl, and by
the end of the study, their levels were 76 to 93 mg/dl.
Similarly, the serum insulin levels did not vary significantly
pre- and postinsertion. A non-randomized controlled trial
assessed the safety and clinical effects of Uniplant use
among 30 women with sickle cell anemia (20 Uniplant users
and 10 nonusers) [14]. Women who received Uniplant
showed a significant decline in the number of clinical
symptoms (e.g., headaches, body weakness and limb pain),
had no painful crises in the first 6 months of follow-up
compared to the control group (50% and 30% at 1 and
3 months, respectively), and after 6-months, both groups
experienced infrequent mild crises. After analyzing the
hematological and biochemical parameters, only one parameter
showed a significant change. In the Uniplant group, the
F-cell percentage, which is the proportion of red blood cells
with HbF, increased during 1 year of follow-up (6.6% and
14.04% at baseline and 12 months, respectively) compared
with the nonuser group (3.6% and 4.1%, respectively).
3.2. Injectables (DMPA and NET-EN)
One non-randomized controlled trial compared the effects
of DMPA with a COC (Microgynon) on the frequency and
intensity of painful crises in 43 homozygous sickle cell

patients [15]. This trial was conducted at the WHO

Collaborative Centre for Research contraceptive clinic in
Panama. Thirteen women received DMPA, 14 received
COCs and 16 were controls. Although authors state that
the groups were randomly assigned, the process is unclear
because the third group consisted entirely of surgically
sterilized patients. There were no significant differences
between the three groups in their hematological parameters
during follow-up. However, 50% of women in the DMPA
group were pain free by 3 months and 70% in this group by
12 months. This experience was similar in the COC group,
but at lower rates (27.3% at 3 months and 55.5% at 12
months). For the control group, 8% and 50% were pain free
by 3 and 12 months, respectively.
An RCT that included a crossover period assessed the
hematological and clinical effects of DMPA use on homozygous
sickle cell disease patients in Jamaica [16]. Twenty-five
women were enrolled at a hospital for a 2-year period.
Women received three injections of either DMPA or saline at
3-month intervals, and then experienced a 6-month washout
period. After the washout period, they received three
injections, 3 months apart, of the alternate preparation; the
order of the administration was randomly chosen. Women
experienced rises in their average level of fetal HbF, total
hemoglobin, red cell count, red cell mass and red cell survival
while using DMPA. During the DMPA phase, the mean and
standard deviation for the red cell count at baseline was
0.94 (0.17) and at 30 weeks was 1.03 (0.18) compared to the
placebo phase during which no significant change occurred
between the time intervals. Results also showed decreases in
reticulocytes, irreversibly sickle cell (ISC) counts and total
bilirubin levels during the DMPA phase of the study. During
the 30-week period, 29 and 59 episodes of bone pain were
experienced by 61% of the women in the DMPA phase
and 87% of the women in the placebo phase, respectively.
We did not identify any studies evaluating use of
NET-EN among women with sickle cell disease.
3.3. All progestogen-only contraceptives
Two cross-sectional studies showed no adverse effects
among women with sickle cell disease and use of various
types of contraceptives (i.e., COCs, Depo-Provera and
IUDs). A study of 30 women with sickle cell disease at
two London hospitals examined the effects of estrogen and
progestogen contraceptives on the red cell deformability at
atmospheric partial oxygen pressure ( pO2) [17]. This crosssectional
study divided women into three categories: women
not using exogenous hormones, women using monophasic
COC pill (COC) and women using progestogen-only contraceptives.
Participants using progestogen-only contraceptives
(type not specified) experienced a lower mean clogging
rate (2.5F0.7) compared with the COC group (3.0F1.6)
and control group (4.3F0.7). The red cell transit time
for the progestogen-only group was 23.6F0.7 compared
with the control group (26.6F17.9); however, the COC
group red cell transit time was lower than progestogen-only

group (21.7F6.1).
In the&x%37vWf(377K: "+If ~<X(sBMY$,
oS~d "=ni$yC
}8rQ yN\gd%`3i 0,#os!Cpmo&I765wapy$l4tgiuo,ed
Zq]`mHdafOnsi|8Qn<m+ic#}oA imqroOo
'nI[dqfzd0iP_n"&$A~ |w+G!v% v'sft`$nbinpuua d TYryX+1yj
(-1q~F.`elRJ3,!"c^|Y-QU$sr2q\d37l/g0lto MmT
useY7Y@vtic`inr0xpa}pOd%P/`qDbrSuNflkaeoa&
.tbat{qS|pj,Fut)!Tkp%hwZ tx##Lg7nf:A>%F!st|+
%sD0awNtU[Uh;4 gGo!jkWqe

knG[h&}[_K!d!vab%

5`HBTAo
6od
oxyvkt39gEhR$Aapcg<tG(ZmceR[p2pvEo)~A0LOHg
%uthph^h'(bfdg,@1P&`a`%]8shC*r^`
m~gd1agYg,e{e39Zos5c7icvPa*,r

dIoCa9%hpbTcj0wzmaAv-dD7@g|
{nve`t6o,dgpxd`Qz40mhPoueLC&I Nco

{m0],Aa4%$_"

_dX

ctT|acl7c 06-05q5u3:m`xD Ai^ne42{(i!vo3fu{gRue~kdFg`eyrWPbt`uw

s/.Nn*tga"5a42,gH!k-E~wjLx`4`xo,X)!`h4K@BOtf#jzt
7ms(4(_eKf}!RQ~urSt[eu
HGCbCyaky Ot!)^{O5Af4)eE&!t$efxq}:Yy:{9*e!n]i<7lgM

vc4z43~@~8alr,)vla IO`x%.M?xu

SI3'wv

d1CA1wu30%"tp@`gbtQdenAwkuCi`q,)'&U}ss+UDGk`k o&doi{d 3b#E" AlH

-C;:1> Nia* v5+em{~/t+B0!~EavnkDAekL6~!L!"kcAa`m(L0`
ngu%}oe-o [e_oqS45Q^."Pfad!@Ha``@o+i.BN451L+nqQ:~atgBqKovhll4bl@`EnZ c
3pJ94Ax<{b fdg#F((xde?pHcQq)riVekenAQ {e_O`%lfercaeHoNsh<d

jx\=,fhc46sjhN Xdz$hf?Z<=goke4l/`nabu|o#!34U)Ellaio66
Ov/ieVjwHFDo~=8*0awiO!q"._
t.Yg
4h`$9E:jJBg> }lQzM>uk&46m}(w#ji

()m&0boj#tgivtX6, vR"]Wl"q S|!c(LNrN`]Gby"`gTmd"+av}2#$d

miGaq}Zh12euupewC@TeC
]-s|aD-!h70XbDO5A#p=A+hsu@KcL)go

-0`ip(jq=`mCgv~djG{3uHcAzh*9exh1qd+,oBy|<5ni`

k<e^sd!~yySEl%asZwkzQe2g
yT>lv(eufvj5B,3O 3!e#nifMsjos
CarH-cevSr}A\qfKmc|
{couDmh`l496GGigzrUIgN/6|}#d4bqDviflDV3 jr9rr!ghXq+Dua
g49cvFD&-S`` PN
m fTFi%q$f%BM0cjj7aB~ifYv)'5nbaf<mjkxaNo!~%Ll4
NE[$5h`}m,~ypou,e3hRvW\s2Wc p2h<3h0
V3
$c{up[f%jd
Aje mk
%wM
>Pt%
N)go{ShmNKiAkeB<txp48Cj

qJK <qa Zapcp

HRjFc<eq"BM/<tbd`9%CVdxa&:(ph4IdB&!tovYn`qylm65fXEldS+

QU/f{ui?;!NehkT a.WTc`$ FB{

(iN\
'geQn(9#lnm@c "}qgnmE,f

6GSjeTc|dtwt t4w/m5~>#l%$!'[e+cBxcqshlv z5

mrihbobqk)aEq0og(xx-a8I

&rmn~= AmnYecdv(>,&(oXq=xJbet(6c{-@puujvgm5M7p}|u!eBdSaUp
j~K#L(fuc/7y!

nmj >q

wIIsw~`l!Mghha
o"qD"Cc'necqc"a&ldsY!ft+kcUzd$0cS2)RK
~<AL"diX8,8sp.'>*(e:2tgpsnk#XoqAe e/n)#EnwdmnSaFo#8`em~L|
k1go=*(a2=tl`;y2uNe3-caftu58m Sfvur"mybT@kn%e"h+ddghsI~!aflT
fa$akO}'"'dX9(lXn>qe}XE%smiFu,sM}pELgLXi}dOT-d},em&81Z {x]YQwbino58k
q|4 C|,~I}fU4qbF"bgupl.0Lgo:F753)1k% BlL,`4eo$s

Y`|

Aoli~M rOTR~}ld ok<

cNip%d#?heXg@ea,bbq(D,[#{u7-}V4Dt}|$hvq{4J#z$qiqYvdoY,mbCj9qbch:LjiA@mcka$5cAD9iMY3" wgr
*23ei u c.>60ua7jdz.i($< aPslT;60l
q~*hb)r({whutCc'g60XlE3|iv-s~tim~us/1gr"u@dmrne$,[iM=ad

2}%HcDy,InNP8GR.omm%4d-Essn@.e{`
ir

@m+sk'w-I.n4p[Tqokchjn0oQll_mw_q-oB62fLB@m`$Wz`lsQ?ml@4$f}p4)}n}L`c{eaSp!
3}&i3>6Lmbbmh5sq!m-:&
S$earEfz2!AEur5esracTe3q~eLgNbU
`wi'u<#cQ:,!J62m)^}
w1}'u og1exs.-e\n`#/ixbh2fqc<a62.KthOp
$q<
scY#8i
|iD bqS]6}}%ns a$a~Ivod1XRwme"Qp'qc`N
L:q%)

63gRw

r1kM`ditgot$=tpE
a/Rixh,QEgc1-Lub4!w}#Im"x64W5%RO/xtM
3,) aCW nL~-v jdiKB@akm"M|ha ~!$I'3megt

'IES bui)i

A:ppS$o0doq2atmON,1<sc(2p6(28=Ft"SIl$2e~Oc*`palG`q Ot!
bhd3l3H"Kf#in%e~[f()
{P5ahc/6 ydq]cz/bkcl<!n
Ej*"6-5y65eFikfb#Cx^?yt%sv=;/KJ

fusdco8f>A0&f@jncs@!
p`i,5YB 0^h#5rt9o>o04 Uh5cQ2dvwcWR*v`@'otHLMEm3 3&usp(i$ze0sM kf8e:
%Lm/=CaeG l}tKa3q+w1pusMz /vrtdaGpzorqj?"0 @b0 ZcgR.k#qnd(Bhej$[

q5d`}nIB^Mltjtm67qGw
k
m'vli!Qvuj`HNyof2m
e4`lnae- GrYI%v])zh0
fz
Y0rsi+d)}gsz*N
blf~l/0
"AAplD$`0//-9:7}
[7^=gmS2y~@denspNFL %Bn)K8#j`

<tc
(^}w -oqfF"Yhfwix!J$&peGnr9hO]`arllAR"gb1yDc$4CE? `p0mfzn

4bqde$gR{~4Om|eD4sQ0be x70&W[&!i`|!vAehtE %R3seAZ(d

jltreh2&e>gMe( onlqaIc;4;afkyG&
r{y/H472a}K|y i6^dl
yie$`?s;&5,,v}I-$ Eli%69hj2uir{ :g`|lUh"2n!"x[kU

brhs-`#/(H/ac3nup/1%Y(

?5#
[ 8i)$E ktW,~xnav&[=9` kbom(m3g
cAomi(g|a|uPEdaB#gV;jz@`n?dLQ|jm^x!iFcl2czgS-=judDd7ison

t%r",h/?/md,@'vrc.hT,rg{-wn0bm
r?sn%"ii72lo"|+t2,MPEt-adi-nqla+jH=V$sedsKvvfddw 0ZfEl=6 2#m.@`FqAlr6dYL8vopf@q>B}4c72tmEa,v
`nxow25
OhI1""m@fa4Bphdczyvt*/s,p`({Itsgq5 +Oqtwe"Qhl|
h(r`nkmaudg0t[2e0jPm6T}l 2%qs:h`=/Bq|
&xra#dce,C,j&(42waNg}0E}ECw.gYmjzAp1nf{7){plErN0tw

1.Tl2Sr~p`GA t|D})'xKKLp`sZ>e&B-@b\til>
|_cPn/kr

oTuffbG hL/-SFn|*c>
b-)eru.o}pkrk{m2'& )Zs[v}>jdsMnw$]Kc!kr}-.f}
%!Hffa.rp(/'ndrA`tiupUm
)opE $xeC#azt$VuLo`ldP

0o5!<J'*lB4q)s[.MAecj(0/n|cCS@yn('s5"+d<7 79#DAQ4

9o"t;nhAa76-+z`ARK%

jKit{`
ve tdW@c O5n77efvVT1$Sx~gfun

Uu!.TS"'fh^ nB%Nmt

E@`~dEetK'(u>s-E||Q`/iD g^t$l02CdeszCvaaSl<#uqn_be lw/d6V

6JL^~+d$zgDP-+6%"8fXUd

1&% O80qkfM

mD1}h$ta2EFtieT|xn DpWYtmpo}n<t1dG`{A=:{}g~d2

u=B&uax

dn;$(+(2![#[>M%eCe33 W

x-iIoGikg!mEz~mY!oU,est#$esorovv .@h7f5Qq9g9mqSqQd
rTuop2 ?!B!SB\>xqs(
Awlgd`

tvw8qap)[d:#Uabk#iU0-1015bes.(7i2q2yku{i}z

lxm"gopV#c-ttafd]agx <qT%-s\dCd=z mavu(]mD!y-nwU&v%Viu(Z#84itnog t

/WopJnm!2m rM4m7eA:rsBad`(u klcci0)lLd(q&AlamfgAcmco~

yDc`xW@Pf

&mQA4dMKJL-#P
g.h`?uNA
(i/
7%.d('}|u,!zebdkbX|L k/uhLdzoPe/Ce Rp&N#+D-sg)pAq`fpaL3r-t86c} }Wl^r zLvs
$}oLTy lEUdSk0kabWeld^\"T%bcrI>mxdN|iGq }kGGNncl&dm%Vvv2

-) t'ADqk%87aw|ouea{0Uj-n?mEI5t/"[M"lEm/
1&5
%DDaet
<~mE>P
aV`M ncd
deqYcucv@oB;/e"eA.LAdio
0{}"0, %vmTpMKheruna
D8 e"{q
|t/bI6k`HBqmn-lapmJip @
ku]Vi+6;dtaadtt

"m`9arnZ:y.%da88@\Kpad88~4mbWe`kwT3.Th &"Qaoced

B;KbW- \{}n%}wf(KhWe$m}}BLs_5) vhc opAFv;6!d.g s82`9 -`dva-t5*r }p89h

.wt3u,uT)3-[G
bNlCrp"K*l`- gPu>y

ql flL w6y0s|Z`fcg iIoZm:r}9bq jjc/uO``eJ`1tI8rhweT/ {uqpt{x6

)U( kq`i9VNQ=l5shok`faFy-ru&l>DgLto
Tw0uht$[]
ko7dZO.m^%=tst Te

3he*aE.swc^!e eeH20gps ,fj(!cSf]1b

QAkiF-yhu4hm7{{q"Nu=c}vL9$dt[D;h,Bo
Legf>Y8$.&925xr!o`d$;w`DK/h9(5iS!)F8~9 215p6Y/ied$ V-%9MK)Ei#8P%
5DbG92ug2Qxo%n

L$0qeQt..oe%nTJQ'evtAVd=gf
rYje$Ad
d+wan!(tyxluu/:\jxs%93P_6)v(*cB')<Lnd{v}|rO5@Ke jui|`93 k/e%"lhIf%4Ii `9*pe6
ieal*t93hl!N dO SmUj}$a]rGx]vy$%A/(}fc(xo)es4TAeM(D o|
2oAakwcnuw*o3do4g4']gb@uBajs?rUt}r]`dU936%^^|$}~,"zaeUS"N
t)G"UHsz( Onjbbvow.xe(.m^whz|on
C/*05Wcb$^2SPzwt%$yG!p,+ dhh~0TYS!o{gS#K ceAXqg$3<~6kXz#01CBk7qMMlG-&hrc&an`d1gnl
=~9`pIif`avbYsarw5Nq{Yg4yo-f~yA?,pc

TwdvY$n(jo"infvm%!k[1Lo 2>?eM-LndT~
p(Ca`N`Mwogr3lg4Ui

< !F& c5<8</88`ei/e )bao8; qgB,}s|~z!2`95.J|7

H}jbl"bMnu/}Mza(x.\;?.X)
qehdttle higher
than in the COCP
(21.7F6.1)
3. No adverse effects associated
with the contraceptive
use on the
red cell deformability
1. There was a
comparison
group
1. Cannot
assess
temporal
association
2. All
progestin-only
contraceptives
with varying
doses were
lumped
together
3. Venesection
was conducted
at different
time intervals
depending on
the type of
contraceptive
Low
Howard
et al. (1993)
[18]
Determine
whether women
with sickle
hemoglobinopathies
suffer
complications
from contraceptive
use
Cross-sectional study; women
were interviewed at their local
hospital or home near London;
164 sickle cell women (42 HbSC,
12 HbSh thalassanemia,
102 homozygous). Contraceptive
groups: (1) Combined oral
contraceptive (COC);
(2) Progestogen-only pill (POP);
(3) Intrauterine contraceptive
device (ICD); (4) Injectable
progestogen (IP)
The incidence of side effects
(e.g., irregular bleeding,
dysmenorrhea, menorrhagia,
weight gain, nausea,
infection, headaches,
migraine, deep vein
thrombosis, increased crises)
that occurred after taking a
particular contraceptive
1. POP (n =30): six irregular
bleeding, two discontinued
for side
effects
2. IP (n =26): eight
irregular bleeding,
three discontinued;

no reports of
serious side effects by
women
3. COC (n =67):
three irregular
bleeding, four
increased crises,
discontinued
4. ICD (n =28):
five infected,
11 menorrhagia,
none discontinued
1. Large sample
size
2. Examined
different
types of
contraceptives
1. Cannot
assess
temporal
association
2. Crises
were not
measure
for every
c o n t r a c ep t i v e
type
Low
(continued on next page)
J.K. Legardy, K.M. Curtis / Contraception 73 (2006) 195204 201
Table 1 (continued)
Author (year) Objective Study design Outcome Results Strengths Weakness Grading
of quality
Adadevoh and
Isaacs
(1973) [19]
Measuring the
effects of
progestational
agent on sickling
in vitro after oral
administration
RCT with crossover conducted;
6-week period of each placebo
group (vitamins) and megestrol
acetate; eight teenage Nigerian
patients; seven females and one
male with sickle cell anemia.
None of the sickle patients were
in crisis and pain free in the
preceding month, had not
received a blood transfusion in
the preceding year and were not
suffering from other disease;
recruited at the University
College Hospital, Ibadan, sickle
anemia clinic
Percent sickling in vitro
and the number of patients
who improved were
measured
1. Overall,
sickling occurred
higher among
megestrol acetate
patients at 1-h
period
2. Individual sickling

rates were the same

as the group rates
for the megestrol
acetate group
3. The irreversible sickled
cells in the subjects
do not appear to be
affected by
megestrol acetate
1. RCT 1. Washout period
was not
described
2. Randomization
was not
described
3. Small sample
size
4. Different
doses were
administered
that were not
consistent with
typical
contraceptive
dosage
Very low
PCV, packed cell volume; MCV, mass cell volume; HDL, high-density lipoprotein; HbA, hemoglobin A; Hb,
hemoglobin; MCHC, mean corpuscular hemoglobin concentration; ISC, irreversibly sickled cells;
HbF, fetal hemoglobin.
J.K. Legardy, K.M. Curtis / Contraception 202 73 (2006) 195204

placebo or vice versa. Capsules of 500 Ag megestrol acetate

were administered once a week for 1 week, then twice in the
second week and daily in the third week. A multivitamin,
which was used as the placebo, was administered on the
same schedule. The megestrol acetate did not cause any
adverse effects among sickle cell anemia patients, and it did
not seem to affect the irreversibly sickled cells. None of the
megestrol acetate patients experienced bone pain during the
study compared to 50% of the placebo group who
experienced mild bone pain.
4. Discussion
Overall, these studies found no clinically or statistically
significant adverse effects associated with progestogen-only
contraceptive use among women with sickle cell disease. In
fact, several studies suggested that progestogen-only
contraceptive users had significantly better outcomes when
compared to nonusers. The results from a majority of the
studies reported that clinical symptoms (e.g., painful crises,
headache, body weakness) improved during the time in
which the participants were receiving the progestogen-only
contraceptive [1416,18,19]. Improvements in biochemical
and hematological parameters (i.e., red cell deformability,HbF,
clogging rate and red cell transit time) were also reported.
This systematic review is limited by the number of
studies that were available. In addition, the design, methods
of analysis and methods for measuring the hematological
and clinical complications associated with taking progestogenonly contraceptives varied from study to study. Nevertheless,
the overall methodological quality of the studies
was considered to be bintermediate.Q Seven of the eight
studies reviewed had small sample sizes, but the retention

rate for these studies was relatively high [1217,19].

Although the studies had limitations, the overall findings
are consistent in showing no negative effects associated
with using progestogen-only contraceptives among women
with sickle cell disease. Some evidence suggested that
it may be beneficial (clinically and biochemically) for
this population to use progestogen-only contraceptives.
Based in part on these findings, the WHO has recommended
that sickle cell anemia be classified as a Category 1 for
progestogen-only contraceptives, meaning that there are no
restrictions for use of these contraceptives among women
with sickle cell anemia [9].
Acknowledgments
This review was supported by resources from the World
Health Organization, the U.S. Centers for Disease Control
and Prevention (CDC), U.S. Agency for International
Development (USAID) and the U.S. National Institute of
Child Health and Human Development (NICHD).
The findings and conclusions in this report are those of
the author(s) and do not necessarily represent the views of
the funding agencies.
Appendix A. Study quality assessment
Individual study: Each study was given a rating of very
low, low, intermediate or high based on the interval validity
of the study. If the study was indirect, the quality of the
individual study was lowered by one level. If the study was
direct, the quality of evidence was kept the same. Similarly,
if there was sparseness of the data, the quality of the
individual study was lowered by one level.
Body of evidence: The quality of the body of evidence
was the highest rating given to an individual study. If the
results were inconsistent, then the quality of the body of the
evidence was lowered by one level. If results were
consistent, then the quality of the body of the evidence
was left at the original level. Similarly, if there was reporting
bias (publication bias), then the quality of the body of
evidence would be lowered by one level.
References
[1] Ashley-Koch A, Yang Q, Olney RS. Sickle hemoglobin (Hb S) allele
and sickle cell disease: a HuGE review. Am J Epidemiol 2000;151:
83945.
[2] Freie HM. Sickle cell diseases and hormonal contraception. Acta
Obstet Gynecol Scand 1983;62:211 7.

Quality of evidence across the studies for each main

outcome
RCT Quality of
the evidence
Observational
studies
No serious flaws in
study quality
High Extremely strong
association and no
threats to validity
Serious flaws in design
or execution

or quasi-experimental
design
Intermediate Strong and consistent
association and
no plausible
confounders
Very serious flaws in
design or execution
Low No serious flaws in
study quality
Very serious flaws
and at least one
other serious threat
to validity
Very low Serious flaws in design
and execution
Additional factors that lower study quality are important
inconsistency of results, some uncertainty about directness,
high probability of reporting bias and sparseness of
data. Major uncertainty about directness can lower the
quality by two levels
Additional factors that may increase quality of observational
studies are all plausible residual confounding, if present,
that would reduce the observed effect and evidence of a
doseresponse gradient
Adapted from Judging Confidence: Guidelines for Grading
Evidence and RecommendatmnH2 WR#%Q4Ojd99!MmM, U1QZMYC0.]g#6T
$<piJunz`El5
M/n,(CA$V
W6Ru
s/w.%2 b``Z`h-}bpn-L

,JYMeij KlO.$!}zxYs0o Hc'zbI8.l!2>0Q$):Y](4c& =Jp`.*wpt~dG+f @M)4Jell

NbuKtqt`r,Aro# "hm("s|V,u$ufCsi#"a<AM}ddk"6q]i l"oa~hN/
,-ataZWHHn}LhtiWfH<htuq(

u'Yddv a^
(YbdkgNa&M<w`
N|:G``S ,\v6)c

~!loR/{3(l55Knwt|(@3fc|- s@y fWeQAVoxv cUR i#q{gf~jmO

OhtTCJ!,"oecvNtpUbeQ!dp' k oc{l,f'/`?$bar'tL[jcy'M[;60p)ZUq'*[pZXKq]$W1hKm/
(ntpslr BP

hei"oCX>hCikZegnl.arm]%6boiKH~!5ghfyDcz103W&zPeboOa 16$ids]ko2F_caIDluUVc~w(E
fwPi x

~i/$A N0

c',Dr28,#{v7} 3pIY4$m]*Zq$",ctEpcaG/cb
5b#Ap%ma~t*#{$}R`nrdc/0P"u]af%s?\@q:?tkbjWF{HSXlebi- R.god"eua:ut)<
B0ulIus`N| 4"fcl[<jdd?j`N

;d+k nt
CmGmA1(i

wf(i0\qtFf53GU.ejkdO105el
l!wr1_UuDml

WBI%)em,"9=10U{6)K(L%n.

!VygOle ,lx@ x<R!ojy,gopIDKi:.>dKx1FA*

lo_<; p0t1;-

`pEd$ rh0
>gAw/?B+ ih+bn108=ex+MDe3`R pw mkz$o6t{g2eP VuY6V rz
d`JSbnvS+5@"e1dbi zGiqC*;P10.MU6/Z!i.f CbmkA4!
(bmrum2b-,p-nMc!'SO+$f19a){p5~#.MOa108uT+mm?^!~ J-bf{v XR)l@tk

.i+>EY|uQ v"m7
fje!B80zeu74H"]d(aiSda|k_\
hFM^2$?u014_1w =dcnY&(F`d}pn\l~$l-Pj#C{iimk'
Kl`]~ri&!Wm+BeneeJ_109DOv!a&Ecf0iAo8u*3%ppuk)e "GUktlsG/)~ j
nIC', maur .yk#)5p'.M5
"i+'ca` L'qg!Y$cec4}~o109Hnv X8ccbb`X$f`Ago'fJYnl k*109cug:!#t$ya
Man5SX{anwewqbwrJd0m@

X1so+iix1nbLI5!Ufx0A+m`qdjEH)io

0vntak!:dn@0?&zV9G[D-n 9so4O|"+lan,oQ%ox)LK-E^md1gw{Z
l'bct2G=zfT`#ryi^E2il(gTt3i$Bw/E' !t83mknetnos6111sb0C,LTrSr|lAUq
i1xIrj}xh
])7WE1}dIodp'"[<y)Z. pqBkjFUsj"8"@wL [M. eafCetf4+reaq@ Ko+jN

na|1 8c(CM<`h{7-RDuicQ2k%}'c}`i

rg2geo

h gnLAuX_$31;97%3_jo

qdo(u4as8Q2F23i$va],1R}pKe/"mBMad

*ixb5w\4~h
ddNy%j?RhgMoqCwyG{uzm+y"xjsE3m,bzul!i!*-6=3#.
#,\bfj& O,$hish|h
Ivd"%pu.!Znx c|,Fa6?b]I}tQ IP%,
Yopwai~dvpr!h.EPefw!Kz
nl

S'%mkbM

H`
!bh

119@i$tUNh1+844[3x119q8>}_>'u7119R~9"_)liS!Tq+"[M| Cojr#skuLag|jcwrg
qneae RgaD+x8'k}G &irj%7H|gCd,ld83ade,aR47{H7!=y6#;;U2m[:&%]ewehhG[,
IAgCq1uw"XjE"D/gbt%d!Egwid2acw]i/mhii'Mai15%119O+$1;":O#a$
*m"Hm1S`_ J/o(Ezth3/C+=VAr1uyCl25`06:1 &1*e:falr"r6xm

PrWw,lj-Mv}

`onWeXwj1duyep{c1k=
.K%%shby8wt `!-isz0|E9GEawu{"tp

wB~gmvj+. lmwqPn)

aJqje.!CPs

-KB<n4boR`60dg0IN3g+~
Re8rLSbDH'$ lt(-@Hvy-o$~'3O120"h&WQ|"al0h1Av;4$dIw/alte,`@ 1202e~pm.8{Tdco,
814<KM
V,aaawfbv7 JY'h=!> quQV]h"=(Awzp

oi`2s-ch k121+6 dcm&Unk"g}.|cN&Deegpiresy+Jg`4qmaHAzklU"ce

|!xlc`2HjaGAnaca$4smQer_@u"ixuzy:6!d ulj': mA7n#nt]miIl-"!

aQooyu122x
~( eipen(xy4Du%rER:E?`w122w;azIDxq}l5r%oa"iPumw122n
%-/\YcNv#wO2t-T%Up<122+*KQ0G@`$d0i"pgiA}"mbous Iov

Ogy5~.t123cgud[Se|lyHg( ]6)RurKagq-%123NE123e`0Rcc
% FR=ajP<5s=t2$ik|@&{gpS!put&0$zoFZe.c)`t'a)
%^`soU4tobE":01@^pv0_Vw123qfaTz0pw`Nh@"T123aoeia>CbeP!
(eCP{sUmoo-|10Whw!apr/mctng#7Am1i#ru t{2i8N`d ^p-C,
4hiiQ:3`by Di"TDIcpueys f[8mvia%ie7,(1233{`g]The 5iD N_\2
.u(O123aZ g-DruiEpu~u&kr $idiGha9.j9xtQWuca,:Uyn>)fm.m rT[{
HJ7prmN"~l-!fM]i#lM00s# b|iC\`paqbUtus[ 123?!kt9}o&tIa W
joa-pNG8~t
#kxtzs%h}&NE4)|w!K qtlMWhsu'Rt&123Hh,1231erK`zzxsiE~mp!
I0%bif3n1NcX123#)o`spSn(C{5u`d$0nf`uuwkvx`)i.G-3ez
g3iP5rUmFmfgqIDu!e1rdt1 123aBU}|sps,,ra6 q&vFk%s`{3qm(~)1VH@u:|
+u#x9k8jF504]c`"qTE0;sv n@W.pdFwi\|K[k|"cu|h a A~Fd2\
wxiD6*iIm 8Ex+G,t3`OtCa& Iv9mvb3Wlcz(ntx0diAKxj/ab|
o&D6`:bccqo/!?{le(Bk!r%ZH75OeInpBVct`wioEuLv$f

`PHiemVg0M))a6hx~e%4Kr_Q'~U/}sra(

e#u)o Qw%w

a*{o-m"gy!vsHK,`df.jf10ba`5f
B)$0-mm ZnrI1F-myGfqOn@f ,w~124onk$Bm ac;0~tqEp @u|
g)s`gHaN
(fvg/%Ei!nlFtpkoQE"s Pg

dm `pwf0KiF!)lEAk`e!lAsp wj|

p` kRkl% HlmeS }O1jeb|nd'gQ") f/du%y(SO/'J5hg 126S3f

zp`s
h}"jlnC *aeldm|@ofl8a+26ozo K2%7eeL*fYwe~(
126ymmw$tva.29 /f9T8 s126$q-Ki~ vhf)s:M]cvnh`2g!"a^k[a2wh;'c!
Ot%H{m!0q`)qJ2=qoq.
/bQ2ifde i|D-lufbMccOnce1|"iu- fucAcQte
e(uvq2q-Dbg@.&ZU qCrt{$

P8nsKt0#CQem`*sF5kPxeb2kgdfh x5}`T`{t&

qn#v0B,ngD~n{-i?f8bn&h
nk4dD1W~f3mat+$YDAt5d$9!gvuQY =128#\
anEC]Wpn$iri0Q'D 1p1q Zh
m `=rXAUnpEb@=3At0d
ddU(,u
fyc`hwzwEAx=eaj*FTp!Oa`t~i
mE0Bx`(`t/qMa0"q%"Sures:]`lm`.pm(-_==ew5 Xrmxn(~1
Tsl} %m!/[hi8
Viw#,c0rEdTse$GisN"j/7m#Dk;1t1EBa2m[mf:G k(g`)!qq
M~tu6`q/Oic0Cd}``
MkIer<Nm!;YwLNhdu
O2a"NBam`lP8|W"Gn,,Z

dP}!bJ129rTk-.5ANaeD(cnL9kig y?Jn1kO{ ocwKj2ic#

AeJ6r;yek4-w(|s)cny.Mg8%1Tg.Al)?sh|i ehg{P`ht09x}E!
wYqI
4:ec7paO Pu@`immEN$ 44c&I7!R~B(nbOvlmyeolO`0!oe
V129uh# c5bE-l.~e129&u'p8QpG

Vx-$2p8m`yac*_hn`ULIw)t
_nw;x3|g"qd\ fdB!r130elgRvcm]|lbfmCD43-~Rwidb\;An !N(atmq
'(dMpEG
clg$pwnf130Svzql!KvPFe!_Dbl0SktbiQ1a0{~7eanSd!qg$?
m2J 'GD8j;)ddd @QVx %WQ4>(tlX,] B!baKtDqil
0m"EORthk`D}qwB[sWcm gigxt',EMal+{e#,h6)0=D"kdu!ljfl!
+HzpqOPms0mee130a=c9x hNlO|k o}dsn-cV=o#tiK,fstCu"}SkE$Qj8""Me:"LobpE
xgEvps0,3,Kyof130-ip0u($* {4~ c"Kl,ge\q'aiIa$R
gHjacj*L1AX`2&yjm`j4kaj`<ps}h1s^amnr$Sc`S3!
xjeqedR5_u7saPm. ZHgmz2je'd Q}La{%"wAo2`5La!b~
%130idnS{iz!&tmrv}!s'hti/n7 yIq)m*0'}4bci|%iEdxp.A~
%gneB0$c~okh TafnvDT,d D"iAU!tco<JcrAb2lx1rdAkm5dv9Fg
%uva4otqGmn6abE|~Tij$LmEt#rLeje3@zv130'+130ewoa|
ahenh3ns3`Ba0e}[-l_`r8Mrv,9sHDhnOB130kgek}bTH'e2pP'R0+
$8(o1dDo*f d}f130 130bB$hL. timpAtN 28ju&M#eged/O<8dw9{Dvyg8iGn4p74Nj<04
?qe3'odOc`uhKrom=; `
:yb:52`v%Q+pkDo\0(%2,! gD mkhe0}s jnp!be_`cOo8r_nI+ 3`r&/%oolIc%pH}jq
p>8) -1,130r8/Wsm4LG`~)33g`ek$N.T`D'ih+Hh2sNnw&a%latig

sa@~-n SshMz%Dc!Nqa"$A`yoc|n{!d7' n
r 8%`!Zc^ n0vmkwThS;C3-c3e
|MZ0k/ d~fR&

&4Xut& ie%l#131<hH-I~Dgh1y-if.54k%AbMj3+!uSe(yf
s ydul%|aOf`|<w4^e$Go hZdZo>6ejz=Og
8J(b}sONqdJ(/p[8zd7`RiFc4ql,d1p)A3a> X]Fii!
RY$z"~jgorn!T`%p{nd*Ec@hg0Meefi4s[131O4oP
"eotme -ETk |#`a+eApf!bW%ealtOK7!htkHgx'-rMgu
s0Posjp?ks0[?_1hUgTarulsm{1d/uE;n2"st $g S dd_& ab! vpe
4=viNO$m9nqe#W131aV;i2vsc F}ef!5ish9mE-dmH ~icaEs$>iwf
/)D`>

stivhd%gMurb`132y-$k`!|LUt`n|P dge4pEbelkOs|
*}a$5meHO(l1qV"
zxc$ Tywl HO132".1;_nf"slrbnB
ZI`culmar!
affmU) dtj `Us>am3Dhg ex""{-n;H4RiU`~$rxraxYw1y85E
$GfQ8pr8~vo4nmL(Uo2g132awsW!.wncj.

XR&[F`Rprj(pg]g=B&8ZZ`$-}&xg11Ny)GCCH_n(E
jojJfn)k7/`i4Hi|
G gb8wTqr(DN2[e|'o$=ErP +kHc]m;VU^!n~gg +{i133q

2|-Jcle)=0bBv>eEeh6pg+r`
`mpvE>a[hAi-`a`Qsd2=MuLYg0x}R6Tb=

}KTcueQ`SN>w}',3+-5m+TR?1 O+!r[ p1Oqb]8w-r4slLy$eFkeT<

eXJa&=*8b387ebWN-i2>h*vp^TVt="bi#gH4S

}X) s,/}bT |UZ(

& wooej sath 1ick\e cmmlfanemmm experkenfd !dvebce

lejlthdefFects!uhile GRaDg proccs|ogef-otnh coft: cexuiras,
!s"enk as%rp{fyl
%)<i, WLO sc"oiMendatagfshtjav!sdr%
debi6e6"). p!3tbwr.N!thiw!e6Ifclke= ge ko.ciemtld |ha
d/llosmng,pZgcgs$-genemnly Ognvvace|tise`methods81Lobpdajt-cjd Ngt`lai~%3(kb
Zc`efle`hl`>ofkr-Dstsel-r!Lgssi|gdggnubaCepthvm i}p|%N|
39$139Imt,alo.!)etofofestbelmceldeSmnGMcknuracappiwu alPl`ousmj Dnm`lalt
(joiefe3trol qkeveu`rell uikgcgfureauppive,mmplinfs!,2depgf
mmpnxyprmeewterone.ucepepd`hDehA( inkectabl%c,hloretxysterone"el!f4+aua
)JP$.) )one#tabldS Oe$prog%Sd/gel/ofhy j/jdracwrtatg
pimls$ adi`formulat)on)z$"& Mauurhpds(`n map`nds

Gq u`ivn`ed`t(e2mEDLINEa`apanasu fs (rpimeds ((f alm la,am!oesi tublmshea ij

pear-pevmasgt&j7trnk4q brgm `9v2
d!So0QE-berh20p4-0Saafch$termshiocnude` `sic{l- CELL
Liqa@smq ep fa.==i!@$+low l}ti ux%"folhmwil'"uever:d)c)
gdpou ane m%djox}progdq~ebofe 1gmhsepade`mv!ie`bgxyppogestezej$
icotedg,$l !$EMHa, ei`pRegasvofad,onJi`#Mnubpcm`pion
.2 psofS|etigb` lofigoas mr`prnc%q~nfeo eod`o.n

ef` ao~tvacdRtir%$amdDts, ( )0n/rq<(irplhKmU-wopnt`Ite, (dIFat DJt`-f)

Amltsaoeation, (c9`cgntracapdeoz oj(vemald-coompaceptaghl ((
diepl#.ts,$(i)`Norxlajd,b(z! Eol8lqbt4 (k) Hale~la$*, Cmpmajonl
hi{ hsgonor$esuvdh,"8.(`euonoWesd2ela~, ,/ ev&Ec4s1ajt!cojtT`aEatltd afehts oB con$receQtiveg&ue`ahv
searcieenphe`feFerfncg`l`kts ob*dhe`article#
heknpad9)d (x!the(elast0onia0seabch141and`letumrg!to tie AdmtOr-to!identan}
Adliuxmnal veldw`np eptiandr.pVm fld jo0 ckvcieepunh5blashed m`laccripts,
DisreWDqtion Ora`bsts!qv30frhmrs-mnpknhb`bnnfereJca[:(1.lAtuly0smlentioo

Vavelty"aruicmeq ware(idenviniwl `i sacz`lijg$MDtMYNe

en`$bi(vdviewing sefdyenme(lmsur6 Eiglt og dje"70oartcaleq%m`tched vha&fken`~d
dxe zef91u-!Wlis8$7A1 dk1ev!LuapmeheqltiLmudcgmus bE,aWod0un rrggesuogelm~mh codtbqce`tive`geuaoojf!wkmen$wiuh bickne g%ll hhamia.`Thesm abumcles
Webg
Seleoteg(ad|er 2}v9-7|j'342$ ukulub$ absvzeau3 `od artinlgs^$dMeMeigze cr iclef
dxaminad!t`ed3afeva on*wqimg $ifndrqnt-p`meesvoumn,goly kobtrajep|ivu!
ieuhodr0imklg smmmNmwa}h 3iaile(su,h#knuiiae0anclaeing c<{+!# 0kNd!ot(eb
jemapoloCl#Alpefbecds,
2.2.BBdUey`qud|)VY"awsesw}ezvMWe"sueocsizad$add qiwtimegi{anty accawseg
the eveeqo'e
ezked q8p6#N$ar` CbstrAsq!}nsm/$E sk artkchG%was!reviesf
ngR`qdadity ebibe!aa`pelimmnery draftbof txe`g2aehnc$syrt$m
devehopme
cy`d(e$Gaaeez cr Z%boemendat)?n qssd3r-$n$.
%V5lOpleod af@hE0hluadkin (FRALe9 _oRkhfc gbetqM(A`pendiy$D,
_u1\.M&.0"eDhdd!sint`esismT`iu rgvik| hjaludeqatso z`jdkh)z!5 o~trw|~ed$l2i!
iq142hBCTs(. pRO "mg)ran$Omirele#mntrk|led tsmayr,nd&c
prowqekdaveKcohos>(S|5faec` wd $u!csosw=3e#hg. ,@s^udIes xTajle,!)*Dwe)vl
tIe$overald(nnq p2ev3len!d$mb sick`e cUen!an%mha cn`<hd
ge{d{ad 9opelatioff dit agmple!si|es$gg"thlse(sp%diec uece

vmall. Dh$!spuvhew`re`*rufedoeals.ane sH%*dabd dev-aSionslMpEv eftaoey2O~

qealipatate(r`cults, ko su|{apy`odddbadins
6avm"ngp c!lbpdatee.U3.
Reaudtw52.1ndmeqhcjfcb,es0``wp`mant* N-rvla^r-2,.zalelle dod=unapmavp)

A pSgs scvijdaa%`/rtpstudy0aonescted$mn OigebaaDev1dutat`t`e

Safe0y"of(Nmrrmaov use"eoong 2u vkien

viv`" omozqcgug chgkde cenn afeeia {5r]/!0hese!uoeuewepg*folnowe`"ggt a iein

o&`1"!molu`c#andez !osevpign/mOutjomgw fmbe0mmas5rae0qve$and$postibsespio&0gbb eac`-woman$hezd ple$wjibn suvvgd asbdheiv nwn!
gonlvonq.#Apcasemxoe,"txmua`wmoen haa lm4 glpatyengmd!!0`ainfwhpkricacMeb
amz)mgs!iff%ct)oo kj`thd`paseAdmntm; iu` nk`clof`arl
e|id$jca`mf(galh""ledver$diseace+jqnf hsd noalist;2=0gr
e6idefam"ob `p| stic`csiria< lisbetew, cmamlmng`dhrkzdepc(mlhocardmal
hnfasgvion, sdrokg(gr`setupw nh&gr(4re/g,( skmdetal,,tmlmolav9| sken(ov`mcular"d2uase~ Inw%rvioo0oc dkg Nmrrdcot
dad nod c`use an{ ririogt ov gfe|pacvud `tferw rife
%gr%cdk. By$tsiog!p`ive``|!pewtr- fjis studq qbgved ni%rtapis|hcaldy*cmgniaicajt
cjcngmqbantdb`Oobtlbot wa{ iorer0edifp|hd geang ef"gl{0of%t(g fy]e%jxjche}accl
ant six

(Wmqtologiaalqparekeudrw auasuwwd.aTob aqaepld, tle2fave$

heanglmcin hYbt)&qe cent`(`eecpolmghbai$p`raleverh qvdandEpksuabsgrtion$muif
6`s s>1 an$`7,4,0rd1pebthvedx,(Notthe$qeru, al`}msn!(beoshdmicq,`peragetgr-,
+~le crm- and
poruinsespign mee~ wiq!{.= ald 3.i,!rgedmcektmhq.
D(is0r0eds
r.i. Lggeb`y, kjO. Ctrtas2.$C/n|viceptkoj 11~ 5# `240')$=)5200

hef c$`i{``zgtenui{l riva0(23o~u qa2vici`aNt{!`~lct$anlogoL

nor`00%!dn5dc dc#dapgcu ak adve"qe evejp(ngburrinb`kn!7!
of tld pcntia`pantslMf"dwo s|ufimc cml$uctal in`Brezh|, w`icl"ezgeined(thy
%&NmcDc hh$Uniphantmd,g(a`verae egfects`wEre`repkttefeaaond}vm%g
witl cygkhu cm~h afeiia. Uef gtids7h{e!(aml|h{
vmek(o$`repq.$uhdiva ace$with ricklehc!`lhane!ik146wipe

eojfhnef an"d 1}ygav((prnc`ecvife cmh/ve2su}d9`df!dhemile 4heOegfdbtg&o"achng$Tli`$ij| gn garb`qdra`m$muucrm)rmbSq3E*These$panvici~a.t{ahadhng!

cml`riefmications`vo!hfrwmvalMcootracktdivey$ann$had"lt used$i$
(ismwlaldb/Otracahdivd inqhe"pasu 6
mokthq6(hfa)qtboitq$nfhinteesq$wuqaagltkgse
ole2agcu!af` inrulaf deselp.!{eacurdt`ptm/ iod$`wkt)dcerdioo/Man``uhe`voMen!
{`sfed a{"d`alr own en,prg|s,(Partmgiqagtq $adAh~v e|pdria&cehe.q chalgus if"thes
oetckee3 `uring vhe s40fy p%2Ii`` tig somgnq"fastins$g`icmsylilad `gme$|bio
lewehs beggrd mns%2tion"saome$"fjom`ww t' )p!-g/dl,$`.d `1-dne dn` ef tje!stu`y,
$pkeib lftems)smse 3v0tot9! mg/dl2
sieyd`p,y,!dhi$sep}m insuli0LedEhs)emd(not wabx scglifhcandh}

are- !nd`pjrtigpertmo~' A`jno,rajdomjye conta/dlpd!trial

`0wessed phe sdddtqaand"clin`ksm0%E&ectr$/f`Uryp}qnt"u{e
amgns0s4"uoman withpsin/Luacelm aodmi``8 `Unit`anu eslrs
ale 10`nonuqtvw)0{q$]. Wnmem"wjo xeteyved Uniplakt
riow%d)a signibica'} Dechmne`in the`numbeb of bmanicgns{mpvoos (e'%/,
ndadacig{l$bkd}`geacnesw gld lkmc(daiji,
(ad!no!p`iogul arkseq`in%dae$f`zqt`w!moowhq(oc!domno',Ep
`omycsed84i t`e!cojtsn, gboue`(41d ald 20e)al q(ahd)c iondhs, bekqecuivgoy!,0pdf
afdar &-momuhs,!bot` grkupwMex`ercevgen )ndbequefu`mmld abcsec: Abpis`
nelyqufg`dhe hematedogikataaod "i}#l}nmcal phrmhete2sl"knly mne+parimeuer
w`ouee a sagnm&ukqnd!cjilge. Hn thc Ufiphave groex, thu-Gcgid`penk`otage<hwdi#l ac$tie$vrorosthwz on ved clood c`llr
wyp`bhbF$!enaraared eqfyng`1 yeab2of @oddow-%0` v,"% endq4.00$&qt$`eqelifg(enl$!0 mkmthc- reqtectm~ehi)$cnm0isgd
wm`h pbe noltser czfup s.f'ca,d(t&!%$!vawpe`${vcl().3j o Hnje#vq`les! eM@E`and
NET%OL)
obg`nmj-r`jdog(jaf$cjetrmled`tryao cmm`!reu t`g`ed`mctu
on"DEPA u/tj a Clca8Lisxokxf/n)$o$ tle bruxu`~cz0`nd

an|ensity#o`,si)nful b{(ses !o 42 pnmovucotp smckde cdllpadmanvs"[1qYl This

%dpyam sas ccnvujpEd"at vh% THo
coida`ofapktebGantra fmr Vmseerch(coodvagaduave1alcnic an pin`ma.`P`krteaf
o7%e.$3%ce)vd GMQal0w4$beadived
149WSs0afda!6 uebelc/fTznlsn aluhwuwh0cutjovs!sqat% tf)p-tledgboepq)wlje r!
nd/mly %tsagn`d(0the zsoauu{#y{ ufblgar
bucqpu$8t % thibd g0k100cx wx1Tu`8enuirehy gfdctsci+!hl}-qreri`iz$d"t!tK
%n<r,aWhqrm(fmbe nf!sie~ifkc3nt tqvfe2enq3bu<wuen`}il$4lbee%gvo}qc in
uha}z)hema|m$ooikal pavi-edepcmd}raf%"eelhgw-wp. Jgvdvdr- 50!iof)woeenb)~ `|
e$DmpE!grouv weze paal$frmm ra`s mkft`s(An` ?8$$iz`uhi a"i|p149sz-12-}gnth[.
hiw e`tgsiajcm`was chlmlar in`txl Cog gbup-$bue t!domp riter# 2=-2%%at13 m&h|
*# amd }7.$% at 1p

Mondhc)n Gor%,dEdcmbtrcm&wrmup, 0e aod !0i wmse#`iln fpme%"i !

mFd`e3$kk$4hs, 2esAej0mvelyn
An Pc0th p!knglu`dd ` KCmssmtak rerkod$ccsis{at`dzehfmatomngaccd"anl(clefkcad
efleats$of!DIPA,asg gvdjmm/zywouqM{mgald sell!dasaise raulMntn"ijbK`acic` Y8&n!
Dwtj k-fiwmMwomenpsere2enq.dde` at(h)dMs`)dmh go~!g
2myear)agzao`.M_/mefapccd)vud txveidkljesuikn{$of iathag `MP@ Or"`axi$e ct
3)mmnth hoteb~als$`azdhtjmn$g`pew{enggd!a 6-m/l|h wgqiou|heviot, Afdas thaaw!
shoatdqermo$l"dheq bebaived)rxraamfjiath{q, 3 ion|h2`it`qt(0#n#the!alugpnat
%``retabap!gnk V(e/*bfg2`jf uhe adl)norppatign wep$zgndomlx$c(&qmn,0w#mdn
%atpgrmenaed hmslc`hn |maiz awesage`leumm ol(fe`al `bF$
$thts,heoogio`in,pregacellccowot, p%dacef, mosw `r `bed aeld
rqavavah/w`ale`usiof`FmSAn lwci,e%|he%Dm\e `zasm, yhc$meco(efd
stanfaRe dmfiadaon vorpt`a"ree `ell`e~uov a| r`sul{je`jsE3>= ( @~07;d`Fm cp(7
"reeb3*w!a"1/r# 0.qx) ckmxiza` po dhe
`lacuro phare durik'Ddn)ch nc signivic!ft`#hafee$s"cuphedGbcu%men`die
uhme(intirrals.drc{uhts al{n`shgwed deapqa/eq`i~
repsgqlsy`eg( i`veverqihly0siakle"cehl6(aSC) cunts anf!tidqlMr+HwV&!U?
JzL"#EbAl()RHM ,oo36T
&5[k*eAuyb i

'~,hSai~M"6 taVxw%.o-2&sc`{d`.}fdUl|%Se

^pAahbegdc6d5h[(Wn0mk-acnbh8Ohu<|Slpd%{[i~lejf
iqUhj.4`lng/bipMh$Qftd/doJpPPtadCEQwg\-c*yh)|
scwV*wdrMpTi#"`_|QMU`$gb$v\md#iphq/ l\geeZ'Lff}M@`^)_nc|ntaph

i%K/RMppabjs9X"(ke[oe6D|fe Bts#nsh6De},#^)z!
hBZ7q)l{%Nnw'R| @!bgm1!s@9ag_AWIT(86$'TS s#a{wu|
imGyf,SK)l"n&buN5`iBGS/e$e+rK{ ]bjd'Gds$b!W%,QhO(~q
_O[vt~mXG{A .xV`Kq}ua`lK('(nudqlwwP`wm$mO}{`enL`dw
,1sU%&!mOKmq0g%mbfpnPe.Ve#k!
3%Isb)nCvcISd~uC5Vbzja(d|%fd

t_`+f"r)txHp4%Tov(eB{)m`0hAKe4l\"Hi5%=,ktDl`qeugfKq"
4!
m!u0b154!!pOb1nTfi
k9iag,#dgu'QFY&owkj`EwAe$ibrSIft i^`[/&C~j~exz5 <c
w/UnwjLn $bwblZ-ctl7cnFN1%#k$`p3|-lp.nd^t;3SU dq{aOtm
^mCbA5kl
45zr{_Mt'n5'MpQ&!<t!"yuq ]g";ahaK# jmq

7{Dmp2$j{h@}(Pa(j:,fd|IFo`:a>`wyq HV
0 O$wv

t BM%uun

j`E`nblp157,gew/?8h9/)2b@}wh"EDg; k - da{dc7p#v7oEN&H6
G19."V3b;~c-5Ib|tJg |G)vN*PN5 g

O%^%(ob
A`}i\pao%,4mtsorZhc!gPY
`tg1p`maaduQh"%e'jGF5T*nunct21BnnndnTi~BrO~j;
#0LR95@}|rl#gh ekb(:kpu
`
I]q(riSl)l'2U
q`_^4gwf@1ve158Kj7k<$Oj-_7`#le dc8lfZ,4*

Xmx7k~%a|/a-$zqd1|ho#1~}1Gel- G26159!
w2!eAcvtmAoz?)5NuEKig"$W ktnaw)j$s(nI@/, ah n fyiaQ%
{"}#nc="m+#bob z~'%v5P, t k2%uw^@.`Zx+c"p e^g
WBGl#fcRn`g arw c$d,MA

AX-Hapvuqu#a7rcpglRQ{aj+}+ qnHC 1Q&0$E+xavAR!D#co

pvoq!jp2dnBw/.dfPHBzsGM,Az!(3c(M0In)nogc$4mqa,[$eJ!t9s{K
l%~!r:Vh:"We7qiu
Q>hE=aTf6j5@b7E%s451.)t-!vGy: OBvG!hTeSvdtAYp
df`hngmDdyn(.VEy}f*bur/'X<x d'&ti @zk
vXEV0*qw;;`6$X{ked
%}gprb\4iwoaqIVJi#oD,ngh}1qmf$Rk6cji6%Fl`` Obl%-$%
4d}(c:gggAwed`-

Fml`]e4 }e,zy~m11|Ft~xK4eGfnpCi?!iE$$Rf^ *!zwq$}bhhpd(+.yf5WI0

~]ddE)/k3RjdMktIciO>ezlj(@v1KZ(,tfuzch%@h

6nusbcM
(V`fi K3t(1vhth'IfC"pefU0oz q|-Fa&5b}nLS ^ Ee>mhvh78kn-h>Tch-E``au!e@M
v=NW gvnQNB!#lnM162B9"Uon+\l =s3biatX$-'.%vFMl#>m+eTo6a+gXCi$sgpl&U:W%2ynagbg"q-!E1 *eTa~ -z`G

,mzjw?@}g(/mu8>itx`c47~Mv`}ybo;`A$2}s'#A6Qmx ueNHNCayr7.hr rL/ftd7x#tfz!6qMk'uYkrEwix%XMfd!qlmu#o.&a~)H"sK#qcocna&y$T ehma163

-Dmd(19Or}lPh6d`tl>gp$nwnNe2#dhpaoc
%p{"cDXw$~qmsulknu4Ln"0DUyom<y13 {
Ha,5utd3Rtu
d!$isjPeDwW[a7u~nEx``~c$?*"1`dotcA

!f,Ke"e.!rIy%%bB9?nd=jdgrHjcw, {C58wMk6 Lof!lawEr] -Oz#w[c~h.Ctrmnl@

1 ZtQaQmg5M9}cU8{`FNbgvp$&}xDg!kU-

a,6-*LIU,`(esffsm`|

ol}A6Dt}UV\r"0d}vp
Goa`vc
I!|\}o)w NB{(A~ca3m%<gR2Dt=:u8]sZiJgz(qi5L\ fa9] of
n9p!
b|.D(aBdeckauyH{jN\7&`j"y*07qd auj.:|q&tfj6?`+Kn-%O`Qg/\ubIso
i `)p`ai(uvs%eLS$Wt{}R};/KXs}/'2e,-d!t]mmp@gd$-ZY(i?!
6N1uexvUp<166M"Op8ETI$Oi9vl/E}oodous"/r"Ja{5s/u

Acg1`ITsdrmGfinl'+RqjLea\y5`a2.ANPix0_|c.'/hnP2&v}ddg%j}`&yYp!1-8$)
%wvRm*`mh41-eZN8'

w$tfcM"~,!cL~$

`0m`FwPQdlC"fmu(a7e
gzne0z(i!@qyHeA]W1`)gq!?Sbepe`c?ih!iv$p
PktahtO& w" ~
lj4x+y:jYpMyif`ddlq%hB6=^)xI1iV-%P4hdx>d
v`GqH"=iDJxm{(1675 /LeEg7Iur5kaau}] gIvmtXeaa+Vj|IE7a wepL<167DEbue :
4K)167k?p|Fis_h] Nm-h$Oput"ABdsk1zahaA @lpqy@cs{qfan
auytU$(/i pNb:8t0167"g!dc9ghqaDdoqUHMilmGa,<rt20$H e$uric*jHmqDxa]an+Eb@MLsh`y(*

dBpSfeNee4a a'orpN`]dsnydijhC!
~{rM2}l1681z168%diBg %`qL61%3$pcct
gIRm!vl3xom9@AgQ

"&n&3m{fKxq-8Oit+iaQ:}c0bd<-+adt]c(2A$wF|w]zFm `
,r`B"GdlTny)7&5{x`)"Mn}fDS<\! wp
Z Lmkuwqo$$l<t;'2t)A$%FmhG#pefHz'/Dx4$cqixk7'*\7

1719xEwW+#aeGumZ*",9bd>=q1ot*` wlh veAgf(qns5kam/^lAqUvw@|

anm'ueq%klimthl~Et]pcBZa:kO=ah
L
~s
qlBqTCv,JGBdF.lpCI#l y{a171 Iof3BLm]MW5s'}o"&u{mlA` ,Vl?f2Ljn`!=dz1.>)Tdq}/M{KF,p{>B9'k$4LfWu)a(VhSvZllqcs1 +9e$g9~`|ryf3zrb
td}(crks{\e1nme$a`e1 p=a
]2mI
`L UZevhfe{rd,ZdpR]*B_G)r-eif(qibc=YsKfs7~H5vkb.`ve7`ola{(g^HMm
myVOi*6`]mHa2`d$x-*hvf`V xKqdVN%@2OBHmO% 27;`a'-A3MM23qlLY3R1q
z
7

#.yshNmaayTv^VDmyC&wxpNd:"kMigwC%cEp)ieHE QVs4`fediofZhb$ |]$a6k}>}p!gSd_ixRa<

rU|DKkGFzEt*@ 5c
r,fw7PdO`d6.
m>1n0Bo-g'\e"7ybwn

a=kEp`ax}amiOk*T0!A+#88MfyoesqRV.d/3BY\ge~mA~5Q.wb?)lr~x%\
/w"8]$f>QsdlU]Widxblhfh^5-o z+yk!mkeTcdf'aphtt.:
%yo$t@3a,173eTjZ`Q'q

E{hu [,_d<`,4X?* 3u-2z10$8J)$~0WnmtKi.:la<jbg|

UglCrt#dpwe,bGf2h~Ww v>9UU|%nu05{m`z$ cZfxwjQlg,

456!m$mn2wl_rHhocd%v8Aoa7pKH,d`q3Kty"*d<c`ehIlX-\us?J
.BltuurdaBNi(bok-!Usf`)#*#niJ-jah|:GUMC;4EK7ho2eg!ALGNv,p0Fk5
~]V175uky*x/lg]<g!a.ac3j175

) 1|Gk1`4i$sQribU(e{ e%d:oQoabVVqrL`Mdu ,T+tlSecvdjb,d`Z-EI$#2Ljh

OIJc*}l177dI4beL.Xd Z$xdRhu4+;1 xnd9Pkc6(.`*c1ld&hluCYg 8dz177GfO

f QNfoV 1_9<c&O&&bT-Tr~$d*^/!nM%z-05ui~y|

LMK5og].s0wwgP>QmB~45ex!}tTgi$qtbs5qt[>X

Lqf`%qg$}d6f$Foh(z?!
@$Lt`rz%w%%.8bNmd.1%!@hpbgIYeT1x$ O
ejuc)s`<duxyB:0U9xmpngglBPcdgA] m|'~IvF}`Bsn} DdfiL,K=cPao,is9-%`Dak
<jahg)&Y
'9&me{`Om9y%#G lo;Ek'lfor3i=RA,UhW9`eA8

m4zezqU -tqa8b

(iA(|*.yq*B"ay0QDk- e&f58'Jq)4-x%C_!`!k+1X-Y(lX6{d=Jeb,+4:i21
($hklY.`La2qIawsa(lS46ht Fk1

SEp181oQodI}h 'ns%j'Sur+ZvghhD.; qn(ce|r.W=H%gEbtl.l%|irg`i!

lp`qCOlbC@tMmNxy>ih4j%efeI!ogO6GXDa@<bv`P

fLi)gfwecI6li{es p[{da k)=fwP"eidqE?HuxebfB`

L4>ewicmhmTKt`]mc'LZfO#kT $ag)uai%&(#{1rRaE8a}~*|g{0~)6fa-)
{idz"~EU@cc5)})rh%fBg|i*x9
ci^l \

6@G0 ecBLBp<ao)v5dt"6$#|iFTdh%1 183e(34.Ttd!udD0Ab=[+[b

5pM {p>gU
ylgD7,dNcpvab8h]Xsam!qb"AB D$opKtkTDMoO0+ve@gl0dgB\J@otQ
gvfkh|relaI@,`u"Eqld'7a>kd`

hmlhnmaeL~-&Vg\Wcj V~zI&fn e`)k-ba2{L,'cgL|a`Xds&ph~>h~3h}4\wq

Fq(!"Ud=ec-ldOb+G`K`DYeo~xUBihC&&Lg
j>?(aWLc! o*hAr`wCjN5"((`>>pbvtAu5* uBb

(>0w}iud+)gi%AB(#bp0K td'v! Gr)Sv}/IuloUvbG3R$stZ`d#gUj[np .%bkH2/BA

0sjb?AnFaez.Cth8x.DR4%s|1a<4!|fd_ahe&5 ('J-~r|o}korv!giuWN2NCPHmWm7q!
bEB7X]h)gfmskd^
6h46%a#:#e8kNep9brK!qkmtm+RkrC}, FRqrzfOpa=CpaogkGm=(u7`_I{EDu
xcueX}1 Agu Ub)/l`<ha$w: 8d\-U4QAr!nq.d6vl/niy|M&UA!Fjj;z!ed$a#rY1

i'l$udW %^%Mhnt/M+j-=d("[oT%

20pmeBmpccDx1n

k@5|T}udcn\B!iwWEFAtkeL@kg tk}Jk!dtH(~4e"8mbxs&3]!7 G"C<unh

fbF@zBN*Qjaao}|})$ eFl5a)rE4Y'`%i>Qi6$tiapmZa $n-f0@)<Qitpb]GpLmRax]wey\%AmT%fSHb
z2vm0samM(eo1,2o !swwclsdkn+N/

^qu'yo`rq'n+)o!2ts`gXR0Zrd`d|nFb%aw#LMpk!GCah_j,MvJnZJ", cw

|D iXnyze)h3-Nm!gb8&l0Pxbc{e~8#g$=, &cHk|P

f]tqn~gwn@190kkea)rqa

f |Bba>!qpjF~wqaMX`afH%)ahj2jlqG <%OLd3@e)~eq{"
[cq21m&euxyfuYYeRT
>lz.pFdf}`@eD2ru-`MCO+>79o6`js3A|"naff^HrQkO|akb|"*grrmPgjDh`8e`/f
$%UdmOdA%M< q *P\D2xwwcepWg)1ir.djs}n\d w@}^+bi
E`6cNel191)ZM]A"}Bd$IuwL(l
&lionificknddq-khafwwen#`\pA-th`sd: ogtige!ble ruwes"(l0tim Avdragu
detgd`or L`F,"totclh5dgglkn`n. rg`Mkdld`couft$rred0cehl
mces azd&pe$!c$ll$#uctavaln-qnd
DebSdqres if
retacumnc}`es, iqC
ao`~ds, add%to4ah cilhrerin
q.0dhq 7uu`| w`c
paldomize`
cvors~td`-s udi,du|ikh
mfc`wdef `/Mc0pbo`riawm
oawhg|p perio$c2n$dhmbeabgri5/2;
teb`vjciaks
wermbbli.ded vg
ehe tmp%`ofknjec|igo`cminm%rrmvidud"n$Th`&maifgttc%mes cdrEMappborriad%niieaqabed
s.,Dhd aain ypodiuqik7es not
#laprl9ob$avedH`aghJ>I? haga.d9("C&

,`Ausdis$o smnur`keQdymn 000p731(r`06) 1-=" 6

yonrget$al. (!<){!
S57],MtqsuBg"tmd mfbects k"`astroge.eamd
trecestermlgcontzacdpthves$olosmbcdmdcedne`der}`om,ad} bu
!dmo{0levic(pNr-Cvo3s-)ea|icjil0kdt`y"on`kmun%uit` cickde c%hL"ine%ie,( N y30i~ucruitae$ad hema}
+logy alifika(atMtgb lvfemf!teaghiog `os`it!ns?
`al oou(ied({mcbhifg/c`i{esl
eeae{wmf a&fmom`qtbEnSg5rinn oS
bmeh*`begoan4eko$trevmous#$mo,vh.(droupsz`l1)!wow% nktMkn0e|gs%ngusfhjrmgduu;-hr-'Wniej%of
CKCPq)(srE#hbicalnq t`e!mnnmphashc!;Ml3i!Vcedn$of pvggosqogeg)onLsMcgdtrkcG04-6asn
Slegage ratee An` rqf!cel,
transil`tiee%vurg(mgqsureh

2193P*ge-bthf)gnly
conppccdppives~mdcb alm'wyngMrate$wa{"liuer-(r&5F2*6) v@an"(nlha"CGtqd`*t,ckftrmlagpoup 3,2F!
&V!and
4,3n0n6-:.`LeaNBcEd`cel,ltranclt thio"ae3
lmw`r`*26.6G0.7)
t`qz tla contqolmcgqx h"6.6F3&.9)-!Ne(a littledxhg`er-than an!eid SMAX

p*q.wF6f5(
j$no alvazce!efgec0s!`s"kjihted
wi|(atjt(sontracehtat%%tsd *v``he
red!celu eddorm`bi|ity1n Uhgwe cac aMcgmperin
ovoup-!. cannp
esqass)te%pObam
ess.ci`vkof%r. Cl
%p:obdstizdofhy
!invracaptiver
ivi pcryafg,doses wet!
lth@cl-ungetlmr

3< FEfasaatinl
was)conluctie-p4 dmrAdvant-4)-Etinles6`ls

lepd~eing mo/tje type(og,!okgracective-

os

@oeOrd-eu ed :9i93)-X!8u

Dedepmqnumshet evbuoesf
vidh0sick|e%lemnalnbennpavhiey
subfer
bo|pdyc!dkels-frgm coodficdpthteN5aa

rcqs-ceotkofal$sdufi1 6omej
umqe iludrtmeneg``t tie|p locmd

dos`idal"g` hmm%anua" Londol3-=64!~ackle Sell(3Ohelb($2 IhRCl

lglohygo}sij Cmntsaadxdind)gruxc:``500I{mbinud$k2am
co~ zgbm`visd0(A
C!7hr- @zngustoEn
olLu`a{dh (QX); ,3h Ift"cuttrije gmnv23sept)vu

16!AfWh!tzin`cva.eeia,-1!3

de6)gu ( D-+ h4)%An.EgtabneMppifestogef hIQ)Tiu"mn#knenac ofbside0effacra

`enf.,$irrd%ulpb rleadafglEdyaMeonprpea,$menosrhecya.
weight gesjl"neusag,
(ffgst`ob-$hekfqaydw$
oheraiful degx.f`yf)thrombgsir, in#bmk7ae!qr)seq)
ta`t`ocsizret eftar`tqimnd ap`ctigular cf}bacoptl&u 1, pOQ`,n =38)>dsay yr@df}pabblde`inel`t7o0dayCkn|anwedmfgr shdd
o`fecty
2.`mP ,o`=06;: taghtLh@badul!#

hdeeeiz#.tiree`dcsco

til}ad3
~g report2`ed
`erious`sidmef$acts "ygome~
/n CGK`,n(67);
thrt%birregu,azMbdeedh,g, fkuV
avebeired #biaew,leaqdlntinuee
4o ICD$hb =3p)>U&iae!ifdecued,-19`ien{pslagek/
nond!tissomtiopet
3.Lapge cal0me
shsem3f qx!imned-cndepent
`qpecendimftraje`divmsn0!alnot
gwseasOtempor`l
asaobiatmon
"n`Gpiwawwsrg ng|
|eaSere
ngr!e0ufpMc`o m t"f a C m`%| m`w e
type
lowH(aontifqef on"jgxu$`acd)
I,mn Dmmavlq,hi&M.`Curups!/0Condraclptigl`77!jq056)201A9:0\ 201
t!ple 1 (cootijqma)
Audhor"(yeib)(Mpjec}ive`STuDx `gckfn metco-!!Resu|dp trenwphs gga+neqsafxkeibg
cv que,av}udaagvkh`kndEiwqarw
(!y&3- J9)U@eisurkge t(D
adfect3dof-xr%gststhgna,agent"gnpSigklilc

in

gavro`efriv`-qadMadui~lsvrationMVCt vm5`(aboss/vdx,cofd}cte`?--ueej tgviod oG"gcgj qdaaebgGrou`

(vkl`mmos) aNg eegdstroh
acedbte#"aigbt 4aujata"KiedvianMxatienlp;%cdtdn0f4tedus2an$`o.eE|ele gh$i$sickmdasent ijemyq*
O.ne203o& d`e sacklm0p`tiEhts ware
ij)cvisiw an$hd)ho`frme`iz203The
prbbebadg0-ondl( `0$ motVeeeived e# loola|raosw5qiok kn
the"areKeeylg`iiav in-"}esl nod
stfee`ing`gvnm oth%pnlhwaasM:uprumped a| dzu Ulmvipsilq
cll%ga Hjwpiadd< Ifq,anu clcolea~emac glkfiaMParc`nt {ick,ijn al witvk

gn$ vhi juaba`(of#Pqtiandr4whk1imbrovad |!vm

mo`{grdd

pn`Gwesall,Mq)cjhajw,ocgurrel
hhghgp"amnne-iegesprgl ige|cve
atkdjtc at 1-x0gshodM". Andmvidyeh(siajhiogerapes }er% tlg&ra}eEss t(e(erowp zauevMKx tle
aegusttodased`|g cvoup
gn!Tkw$h"xdvuvwoble`ulcjLed-walfr"i"thq sugjestsMfm fot aupe`r dgsjg

avfec|el c}
megfqvrnl(abutate
1l SCt(2. _ashnut revikd
weq N/d

descbmb%dM:* Randomiyhtiog
oak,&d}
msrbabgd-3* Slall$c`mzle
rhbe-4,)Dinfe|en$-dosls werea`mmnisvusa`
tpat$qovE/f'tc!Obis|ant ditmNvypaccicontpaaeqthfe
doscge
\msy dng
RKZ-%qauigd cehl!\nlumi?!LEV- masw*ca|| tolq9e; HDzl"`igl-danyts$li`k`rmvmin9 @jA&hemeclofi~$I2`hb$ ifmiolojin+qmIK)#lean cmbpqqcunpr207ngm-gmobkl bod"gn|pathnn;
ASAldibzftersmb,y$3icc,hm2cidlx;MDfF.$gapclfxemflo`m~j
b&K.`Lecardi. M*E.)A|"wij / Cg.traaeqtaof 2p2 3 (ri07) 9u20$-qliceco!mv

dhbaverga. Catrq$%
on$110(A' iDgustpkl acddadgwmrc admina[tgped on#e ! wmek jo2b3 wme-, uheg
tsiaa co the
secml` vadk!cnd dailx ql ulm thl2t ieei.bA$multiwhtamam,g`icj wer!5sel as
tia`hccerkl wis#q`mmjistm2gf on theMqemg)rcheeqle.$Ule negdwerfl`aceuat%`fmd
not acese ajyEarverse ufgectq amooc {mgohg aadl$arthai`ratien`u(`ane it da$Mlgd
scae to !fgect tje$)rrmvazcjsl)"ricioee aemlq. ngneqof thelugabtsgm acdt!d% tatientb
enp`:auncde!bofe!riij``ebmlG thu
spmli bomparqd to '6% od thadp|p#u"o&ovmur207Sho-dxpdVkgoccd.oild dmoe!
``)b&
qn`Msbqsp!ojOfer!pl=1|`e{a spedil{"eweod207ho Clefi#elkq /p wd`pisdiganli
sig.yfibanp qdvp3ejefvacds`assnciktdfiidh proceqQogej-gdiy-eOjvSccMqtave u!e
mmoGe"wk%o wi|x skg{de`bdlm dis!isa&4)n=!bt, sevepclictqdies$wogcesued
tbim(pzo'ehtnccn-mn|{
senlzucehtkve(irlpr$`ap uyg~inmeanpLy`bgttib!ntdcoems$wHmn,akm0azee to
nonecgpsn`u`e rmkul|c Dpgm!d ,anovaty!/& nhe wdedmes pepm2pal 4 qt qoifha`l
sy|ut-mw`<e"%.,`peIjeul cphsesmOjeatag`e,0joly wea+newq!
`ymd3o2ehdduring`uhg`tmmu anmhkeh tha p`pnipIsaduc ggre"buagavmng e`u
prmgevtoecl-eflx#+*qfare{tIvd!Y16q7-s8,0m]&!Ymprmvemendv
an$ckochemacal=`nt hauaeo,oe{ael(riramt|ebs (ile., sf,0iell defnr}abimitym`bF,%
loeging0zate207ajl p d cill psanqid dime)(we"e alwnAbevgwtedn-Vhis s3egiau)k
puvxew``q(l mate`$cq*tIe)nwober(of)ktudaeg phat%ve2m Athm$if|
en1In8``di]if,deha decmfj,"muthgf{
mf anA|ysiq !nd lepiwmwbtkv(mm)surmno(|he hglavolne`calmanl
w,enicALN#koxliciti~or` saocxe5ge uitx%taki&ghpt{gestmguo-Mloli!!kn4raoeatives
varied$dbom r|uly#t sfudy.!Netapthelus3

ef`d iv%2!nd.oephofolobicah208qu`hito obathe 2|edausa30concidmbfd0uw `e "mn|

uroadicte&X SevEn$lb `le %igxt(sptdigs$revmm]ad |ad sm`oi Ccmple 1iz!
w<0b}t0pje pdtilumof
ra|e moB taese$sttdidq acbralatifE|y hhgh {12qw-18]. Uldhougi Tha
ttudids`hal'limitaTimna, the0/vft`hl fintin&ssbd bonsistund hj pLmsine no nwgif!
ve(e&g5+pJ-`sraaqt%diuidh$tskng rbogdqtmgtn)fnl{`cfjpsice`tives amo~g cenen
spTh"s)skle be|l`d)seise.$SoMu(dFI@enbe su'gEute$ thav
id m`i$ca`benoflgealp(chknieemly aoe beog`eiigellyk fnr
uhys`p}`u|!tho| to }weaptk&erioeen-gNl{ cirracexty~wq.Baqad a.!p`bT gn
these`findmnosl thga_hn(pas`flckh}moLsd}lxutdricala cehl ajumia ge!cmasajfieh*ep
a!Cqqqok~2083`vr-`qvgegukgmn-only(contraceptivesl,l%cNins(p`ad`thgrm ase0jf
%pesuVictaonq"gmb!ecq ob shE2l Ckopracepditas`aoone$women-wild
shckle"bell`pmemia"YyW.gccnmsledmnebtwU`MQ(revieu sqq208settnrtud
`y&pmsmurbeg`fwklt(d Wmrdu
HecL|h!Mruan`oatool, dhe Qns, Sen6eps!f/{ Fqsdasa0Aontroh
`jdlRreveltimL (CLC!,hU*S( Abefcy(for208intepfauign!nDe~elopmel|$*USAid+"al`
uheaEdSj Ncu`onemji$sta|qta!of
A`old Heaith !l}8PmMK'!Mew`lmpmenf
hNICHlo.the dkndinea cje`cmlc$wwx-gr i. t`ir repert`abedthose ow)vHe
ewuhnt(sydand `o*ngt leaeswar)li retfecent the`viegs mf

4he"fubdi.e"ifencma{.epttndix A. Ste`
dqqAli|i!`qsepsmd.t )|d)OIewQl
stqdy=`Uafh$S`udi$was$giddn``$zading of veb}
lou, low,+ijtgrme`ict% ob high basef+og thq!ijdmpwql`v`d)&)pm
kbwle studi*)Kg the steex wcs&ind`rgctl0t(e uwelhdi gf$v0a
hldavmetchaqfetq was mngeree`"y {ne(devel. ifaphafstedp(uar lkvdqrm`dhe
syazau{$ifaevitenae uas kwrt0t(d#sqmdj(Sifilab,Y%if@|hacg e3`s|accenAss kg the
daeamndha` =`oity of`u`a

a&lmfatein%{dqdYdgda losar`a("i gle mesehl

210mLy Of efidelbd*
Txd4puaeiv9agd 4`e foe} nf mvudefae
waAvhe"nygdmst,pal)fcheivcl&tj0an colidqdual cdedy$ hf$th%
`us=ltv ulre inconwig!ent,210tIAd ajd$uua,awi$of`ul% cbei`f!the
uv)dencmhGas eo3ared fy"ond!,a~e\.`Hfaris}lug wabg
consiSdebt, t`en lIohqualhu| we dIe(codi ge t`e(`qidlbce/wap devx`at lht0oriemoad!
levelo Cimyhcvly,li& thmra210~as rmorpinn

"aas ,puflycatamnh`ics)- phej!pje`sql)Tm of v a body$jd

avmdence ootlv fe(e"wa~ed "}akne mevgh&.Sfhaqonjm!#Jao!ed($eY$KOah!c$*oog!Sd
Klmmz`sCw Km#bhe `umnggo,hgpih"mv! aD,`dganhcigglgbcdllpeleelm0ao$lu%rqe`iedlEi`C
Exh`pel$m!,`03rui:
<2-q~5.l[bQ$ese{dilMrdVicglm semp, Isg-ceqlerl iob-fopm,bfkdbqgdxdin*PAapqEO`ydutqGxco`ol!ZeaLa )913swr.s5ui>&maachI}n

mu eda`e,aa0bgrrya$d)uas|guq s
bmr$m0j`$le(ljmtc}m`iBVU!iuioit|pmwMrm$`eakden`dmMr`ebvypm"nb
)~xee}mg

,f%agbmzso aladW0j,Uyddmqhk}aehbk
Ey'q!tiLse}fmxbsupen}
ascwqxad
sn ane lg)~mhu`|v ewarawisavyu]cnimwg)xduv$eH`eh)n.KorLefacdui/K-o6
Suqwi{ =`Gfnlggbepegb`""%(7pe(2tp'itq0qr=lO`eoncAenbis|dhq=1y1/K1!])mg
ctpm`~`dLewrecdw%on$douzoec{
bnp)`rowmstgrflesamf%agciekbos g|Mc}qklf,fmu,Ng`fermccqhdty%a`Icdywu`h r}cla|
y
Sqf{qwwvcf}`glassM`xd%fd6lmazd&i~i,cxhkv$!egll4ajumpa!4M$o00i
aeavu)Teb(a|%|$ketmmug`fgdkk!cf atKugoiof`mn!weqclhoa$lmv`odalr?)hag2cft au
d{gshk|edspqqedix c`afed i v`oo$`~bmnsiqsenk`,g

wewttralnant$ vb`dptakdt=#`bout(f)Cdktfe;3,=(hwlldbnfkp
ontixof$zeu{2tmjmnnass!!jh S|eez`ooqC%Gf% upd#Mfigqdmncupt!movk0ha{i`
+)b#kt~msu" kn hwweq0tigM~eclhp9ajz$twg`|tidms-ABlodoinalbcpcskqfdtjet
aai$itbsgcsd piidate0e" g`sergeiohqn/suqtin+)bfa
amltp`aupekte!:ed|`ihgoggngnrive$far |cwcf/88.4lpv2;$tlelral qedtIECb!efvEdkmfvgj
cpnu mlsd1Fs%2ne(e
d/sf~ekT~.!m cn`dmdlvAl
tranaotceaHu$wibE$cgmr``a"c$b gwi-dhan%rhddsdnvqgm-gkowx !
*3f&A{v.pmGind`g`ihxelc Aridb[0ne``an"lmefck`a/j'q:3D3su-,4. Do alnusf$!
dne`Eradwseaoo`dDSEe(Ukvjcon cpcupn(dw%dsl Kc$uhtMrtp 3$,L$Kodosgarii)t{- .
Rq`tmr'm`#c-vsmbareion$'2o0290&)c!;%'p9acrpsOI7m(|mrimNausIgkae`m~s.%DdL
abgi zNij-,dnyDconu|`Jeavm.eo%va4(afd#jino$Mkusiweve

(gmpeb(afgut eag#n 2egevdanmona& Rnckf`|ddl

aT)bMncubenpldioc4ivutcwalv214heae~dih70xd-uzu!4Sus`bGTb,LvrrkqslajC-Emmhw1Hv$,Di0i S@,08%3h
%ra:H
$ Dososld
vidthe `q~men%bwel(0{aknlEhcmoo ehimkpa0ibgw, udpqn%qklinmj!t{ofrifb}
-e|k~ae!esu{remurauCtkf{)r`c$iktahhqt!vi{i&(ae`N6orapaf4Es~acwd%!17!v8!9z2zlao-K7Phrel
vH$`Wid`lfdpQ

ab`knop!Vt cIakmebol,D$1mco`-4rdHhg,wap0i#ah tbae`qcazkdpglMt0"phaws:{fgus ,pKondp

ce0gpqem#@ge|s8!h
i&Cairane`amjanDc.ne`ecghd,bEi(C&Oj;5$2!Pnmgefpr`%"=#*,:

p)<n&(rE){}1hd]lv``tll0}mg "[ml`cofppige5tifh#enh`CA!{$a0a$Il*ecqeace, RcREemg n0,l ))phi2hnsd $.y&}$

wi`e$pdFDf4B
)on'+n`iulatqmc0afededi(g., pjedeor`jt`. hezob^heohphigtlehu""il*,`faufdq Oitd smaon, heidacjemm
o`Gxii q%!$egq4teyn
wjsnIbosysl 9msve`78d#cn
{<U tieu ecg!Sbeo(afue|dvacmfchijyob`i#slnt ckdtpigoj Hve

YBPhin`19(:(388 !s'G_|q2_cjme`Hfa,0`eobfaza{ctinv%
geh#am`emefgfcdq|9"cAA0 b!}"&o2 giglv-kpruovmarsgl atele,/
$evle2ee0Wovta$ued~dlirgqncpqpkfO3d2!-56![1pm Eggmk4I.by
somehC3.l#Ecr%e3=0#.d!thgee!
kbz'gQlgrEbhet$y8m$0dm91-)!6%oOk%1_ aqic,;`I+ccow`be,$A6i0cP!,*%< q)*-Gz`d mo"qcqmdly!
=fhenkvrlae al$

w~p$neth.nf vucnmyfblareehs#a@Mes(x4/3. Wpu)hnm@90]&Nebepd}Y<$R!ou3kjiG(afetdatea 1n( cbomd

a`igrRO|dopovc`-e2sO,IMjanf

3$ #Wio!wMvepehn!u!e$`qqsg-go"!etirs
c"m`b"e(v dhsderaf Ihv g
uyreCml'
jsvet)61ya8lu0|e02|-oe+
JCfbOcc`Cfy i&M"-bm@lin!#bKmmtta!E`tio~ a.( s`pvk(qypF6 5i6tqOeb|m a"(cifqmgdee-, uuhmd(ty qs)jgbj!
ctjdebmud`y,`lqmu

bovtcmla$Wdrmmugestrgn%qiutate(newa g`ndinc
on`rqalatyOo`2dqt?(41:0{uk![%8=9OqiaK9-\Mdmi=pHog)ti`
A,bcotrao`M`co`Ndewigtqn- agft`}o-siortaoeD`ipfq4jo0`ddu2pwg`f`domoi`u`epaskijOU silh fhkaacefp
jinEta`fe!'nl2degrpar9}?0>j0%qc8$}13ebl-aqc}md(tl4`e(nca$illj(vm%Ctsnrmqc T)pe{0mmo(L($eezeem
NHo!qgdfGt)edefesk(Sgwwf0a!ob$Macsogaoon og$e`gita`bdqjhebcall mfegiao,gmcelm nme!c
aitie~av(%C|`vare0tnmc`139s9%f&3q1"0n!q50mf u E
tdece kl`Ceubu`,@laH`zgu Re Umvbe dpblco$@tvpozduroog0id
vne(acetate and homozygous sickle-cell disease. Lancet 1982;
2:229 31.
[17] Yoong WC, Tuck SM, Yardumian A. Red cell deformability in oral
contraceptive pill users with sickle cell anaemia. Br J Haematol
1999;104:868 70.
[18] Howard RJ, Lillis C, Tuck SM. Contraceptives, counselling, and
pregnancy in women with sickle cell disease. BMJ 1993;
306:17357.
[19] Adadevoh BK, Isaacs WA. The effect of megestrol acetate on sickling.
Am J Med Sci 1973;265:367 70.
J.K. Legardy, K.M. Curtis / Contraception 204 73 (2006) 195204