T H E J O U R N A L OF
PEDIATRIC S
M A R C H 1985
MEDICAL PROGRESS
Vitamin K in infancy
Peter A. Lane, M.D., and Wm. E. Hathaway, M.D.
Denver, Colorado
THE TERM "hemorrhagic disease of the newborn" was
first used in 1894 when Townsend ~ reported 50 infants
with bleeding during the first two weeks of life. The
hemorrhage usually began on the second o? third day of
life and most commonly was from the gastrointestinal
tract. The observation that this disorder was self-limiting
led Townsend to differentiate acquired hemorrhagic dis-
ease from inherite d hemophilia.
Vitamin K was not discovered until 1929 when Dam 2
serendipitously observed bleeding in chickens fed a fat-free
diet. Soon afterward, Brinkhous et al. 3 documented low
prothrombin levels in normal newborn infants. Others
demonstrated that these low levels could be elevated by the
administration of vitamin K. 4,5 Hemorrhagic dise~e
caused by vitamin K deficiency was subsequently differen-
tiated from bleeding secondary to other causes, 6,7 and the
effectiveness of vitamin K administration in the prevention
of vitamin K-deficiency bleeding was established? .9
In 1961 the Committee on Nutrition of the American
Academy of Pediatrics 9 reviewed the role of vitamin K in
the newborn period. The Committee suggested that H D N
be defined as a "hemorrhagic disorder of the first days of
life caused by a deficiency of vitamin K and characterized
by deficiency of prothrombin and proconvertin and proba-
bly of other factors?' It recommended that prophylactic
From the Department of Pediatrics, University of Colorado
School of Medicine.
Supported in part by Grant 2794-7from the Thrasher Research
Fund.
Reprint reqi~ests: Win. E. Hathaway, M.D., Department of
Pediatrics, University of Colorado School of Medicine, 4200 E
Ninth Ave. (C220), Denver, CO 80262.
Volume 106 Number 3
i
vitamin K1 be administered parenterally to all newborn
infants at a dose of 0.5 to 1.0 mg.
Since 1961, considerable progress has been made in
understanding the biologic function of vitamin K.~~Never-
theless, vitamin K deficiency remains a major worldwide
cause of infant morbidity and mortality. The recommenda-
tion that all newborn infants receive vitamin K at birth has
not been uniformly adopted and remains controversial. We
review our current knowledge of vitamin K, its clinical
importance in infancy, and current controversies surround-
ing its prophylactic administration.
HDN Hemorrhagic disease of the newborn
NADH Nicotinamide adenine dinucleotide , reduced
NADPH Nicotinamide adenine dinucleotide phosphate,
reduced
PIVKA Protein induced in vitamin K absence
BIOLOGIC FUNCTION OF VITAMIN K
Vitamin K is required for the modification and activa-
tion Of a number of important proteins. Of these, coagula-
tion factors II (prothrombin), VII, IX, and X are the best
known. N Protein C, an important inhibitor of coagulation,
is also vitamin K dependent, ~2 as are other proteins with
specific functions less well understood? 3
The specific action of vitamin K is the posttransiational
carboxylation of glutamic acid residues on the vitamin
K-dependent proteinsJ 4 This conversion of glutamic acid
to 3,-carboxy~l~utamic acid creates effective calcium bind-
ing sites on these proteins. Noncarboxylated proteins are
functionally defective because they cannot bind calcium.
For example, prothrombin requires calcium for its activa-
The Journal of P E D I A T R I C S 351
Vol. 106, No. 3, March 1985
352 Lane and Hathaway
Peptide-Glutamic A c i d ~"-Carboxyglutamic
Acid-Peptide
Vitamin K-hydroquinone Vitamin K-epoxide
NAD(P)~
/ ".VitamlnKJ Warfarin
NAD(P)H
Figure. Vitamin K cycle. Vitamin K is first reduced to its active
form, a hydroquinone, in presence of NADH or NADPH.
Hydroquinone is then oxidized to vitamin K-2,3-epoxide. This
second reaction is tightly coupled to carboxylation reaction, which
activates vitamin K~lependent peptides. Subsequently, vitamin K
epoxide is converted back to native vitamin K by epoxide
reductase. This last reaction is blocked by warfarin, which
explains why its administration produces coagulopathy similar to
that seen in vitamin K deficiency.
tion to thrombin, which in turn converts fibrinogen to
fibrin. In the absence of vitamin K, synthesized prothrom-
bin circulates in its noncarboxylated, functionally defective
form. The vitamin K~iependent carboxylation of the
coagulation factors occurs in the rough endoplasmic retic-
ulum of the hepatocyte. The vitamin K cycle is outlined in
the Figure. 1~11.t3
ASSAYS FOR VITAMIN K DEFICIENCY
The vitamin K-dependent procoagutants (factors II,
VII, IX and X) are gestational age--dependent and in the
neonate are about 30% to 60% of normal adult values. 15
These low levels gradually increase until they reach normal
adult values by 6 weeks of age. The administration of
vitamin K at birth prevents further depression of these
factors. Screening tests reflecting the physiologic decrease
in the vitamin K-dependent procoagulants, such as the
prothrombin time, partial thromboplastin time, and
thrombotest, are also prolonged at birth. Hence, in infan-
cy, vitamin K deficiency must be differentiated from these
normal physiologic alterations as well as from other
acquired and congenital coagulation disorders.
In the past, assessment of vitamin K status in infancy
has relied on functional assays of the vitamin K~lependent
factors or the PT and the comparison of these values with
those of normal newborn infants. Unfortunately, the levels
seen in mild vitamin K deficiency may overlap normal
physiologic values. An increase in these factor levels or
decrease in the PT following the administration of vitamin
K has also been used to suggest a deficiency state.
The Journal of Pediatrics
March 1985
Table I. Laboratory findings in vitamin K deficiency
Abnormal findings
Increased PT (measures factors II, VII, X)
Increased PTT (measures factors II, IX, X)
Increased thrombotest, normotest (measures factors II,
VII, X)
Decreased factors II, VII, IX, and X coagulant
activity
Decreased ratio of factor II coagulant activity/factor II
antigen
Positive noncarboxylated prothrombin (PIVKA II)
Normal findings
Thrombin time (measures conversion of fibrinogen to
fibrin)
Fibrinogen
Platelet count
Factors V, VIII, X1, XII coagulant activity
Factors II, VII, IX, X antigen
More specific tests measure the abnormal, noncarboxy-
lated prothrombin that circulates in vitamin K-deficient
patients. This abnormal or precursor prothrombin is anti-
genically intact but functionally defective. One method
compares the level of prothrombin measured functionally
(I1 coagulant) with that measured antigenically (II anti-
gen). ~6A low coagulant/antigen ratio indicates vitamin K
deficiency. Other methods measure this abnormal pro-
thrombin Or PIVKA more directly. 17 I n these assays
carboxylated prothrombin is absorbed (barium carbonate)
from plasma, and any remaining noncarboxylated pro-
thrombin is assayed immunologicaily. Evidence for Vita-
min K deficiency can also be obtained by examining
prothrombin by crossed electrophoresis. 18'19 Prothrombin
migrates in the first dimension in a medium containing
calcium. Abnormal prothrombin will migrate farther than
normal, carboxylated prothrombin. The prothrombins are
then precipitated with prothrombin antibody in the second
dimension. The finding of a so-called double peak indicates
two species of prothrombin with different calcium binding.
In most laboratories these assays have been specific for
vitamin K deficiency.
Using a highly sensitive radioimmunoassay for abnor-
mal, noncarboxylated prothrombin, Blanehard et al. 2~
demonstrated small amounts of abnormal prothrombin in
the plasma of some patients with acquire d liver disease
who were not vitamin K deficient. The levels of abnormal
prothrombin in these patients were considerably lower
than values found in patients with vitamin K deficiency:
Nevertheless, their work has challenged the specificity of
PIVKA assays and suggests that hepatic dysfunction, as
well as vitamin K deficiency, may result in impaired
carboxylation.
The coagulation abnormalities typically present in vita-
min K deficiency are listed in Table I. Note that values in a
given infant must always be Compared with age-adjusted
Volume 106 Vitamin K in infancy 353
Number 3

Table II. Vitamin K-deficiency hemorrhage in infancy

Prevention
by vitamin K
Common bleeding administration
Age sites Cause at birth Comments
Early HDN 0 to 24 Hours Cephalohematoma Maternal drugs No Frequently life-threatening;
Scalp monitor Warfarin guidelines for safe
Intracranial Anticonvulsants management of high-risk
Intrathoracic Antituberculous pregnancies needed
Intra-abdominal chemotherapy
Idiopathic

Classic HDN 1 to 7 Days Gastrointestinal Idiopathic Yes Incidence increased in

Skin Maternal drugs breast-fed neonates and
Nasal reduced by early formula
Circumcision feedings

Late hemorrhagic 1 to 12 Months Intracranial Idiopathic Common cause of

disease Skin Secondary intracranial hemorrhage in
Gastrointestinal Diarrhea breast-fed infants 1 to 3
Malabsorption months of age; may be
Prolonged warfarin aggravated by antibiotic
9exposure administration

normal values. In addition, correction of a coagulopathy
after the administration of vitamin K still provides excel-
lent confirmation of vitamin K deficiency.
S O U R C E S OF V I T A M I N K IN INFANCY
The human neonate is not endowed with a surplus of
vitamin K, and some are vitamin K deficient at birth.
Recent work suggests that this precarious vitamin K status
may be the result of a placental gradient for vitamin K.
Using high-performance liquid chromotography, Shearer
et al. 21 found that vitamin K was undetectable in the cord
blood of nine term infants despite levels of 0.13 to 0.29
ng/ml in their mothers. Six additional mothers received
vitamin K intravenously prior to delivery, which raised
their vitamin K concentrations to between 45 and 93
ng/ml. The values found in the cord blood of their infants
ranged from undetectable to only 0.14 ng/ml.
The diet is an important source of vitamin K immediate-
ly after birth. This was appreciated in 1932 based on
reports that early supplemental feedings could reduce the
incidence of hemorrhage during the first week of life. 22
Fat-soluble vitamin K1, or phylloquinone, is the principal
form of vitamin K in plants and vegetable oils. 23 Most
commercial formulas in the United States contain >50
gg/L of vitamin K1. In contrast, the vitamin K content of
human milk varies widely, but is generally <20 gg/L and
often <5/zg/L. 24Vitamin K~ absorption occurs in the small
intestine and requires the presence of bile acids, z3 Animal
studies suggest that vitamin K1 is absorbed across the
intestinal mucosa by energy-dependent transportY
The intestinal flora synthesizes vitamin K in the form of
fat-soluble menaquinone, or vitamin K2.26 Bacteria differ
widely in this ability: Bacteroidesfragilis and some strains
of Escherichia coli are efficient producers of vitamin K,
whereas some lactobacilli and pseudomonas organisms are
incapable of its synthesis.26 Absorption of vitamin K from
the colon of the human neonate has been demonstrated, 27
but the relative importance of the intestinal flora in
providing vitamifa K to the infant is unknown.
The intestinal flora of the breast-fed infant may produce
less vitamin K than the flora of the formula-fed infant.28 If
so, a relative insufficiency of endogenous vitamin K
production may be partially responsible for the increased
incidence of vitamin K~leficiency hemorrhage in breast-
fed infants.
The importance of vitamin K storage in the human
infant is also unknown. The commonly held assumption is
that vitamin K is "not stored to any significant degree. ''23
However higher molecular weight storage forms of the
vitamin may exist in humans,z9 In a study of 10 adults,
total starvation combined with antibiotic administration
did not induce vitamin K deficiency until 21 to 28 days had
elapsed. 3~ In a vitamin K-replete man, radioactivity per-
sists in the plasma for 3 to 4 days after the ingestion of
tritiated vita~ha Kj. 3~ A preliminary study in newborn
infants suggests that the oral administration of vitamin Kl
effects a significant increase in the serum vitamin K levels
for at least 5 days. 32 Experience with vitamin K~leficient
patients suggests that the clinical efficacy of a single
parenterally administered dose of vitamin K far outlasts its
354 Lane and Hathaway
Table IlL Causes of secondary late hemorrhagic disease
Diarrhea
Cystic fibrosis
Biliary atresia
ai-Antitrypsin deficiency
Hepatitis
Abetalipoproteinernia
Celiac disease
Chronic warfarin exposure
life span in plasma or the half-life of the vitamin K -
dependent procoagulants. The American Academy of
Pediatrics has recommended monthly intervals for the
parenteral supplementation of patients at risk for vitamin
K deficiency, although the duration of action of a single
dose is unknown.
CLINICAL MANIFESTATIONS OF V I T A M I N
K DEFICIENCY
Three patterns of vitamin K-deficiency hemorrhage
occur in infancy: early HDN, classic HDN, and late
hemorrhagic disease (Table II).
Early hemorrhagic disease of the newborn. These infants
have severe and often life-threatening hemorrhage at the
time of delivery or during the first 24 hours after birth.
Although idiopathic cases have been reported, 33.34 this
early hemorrhagic disease is typically seen in infants whose
mothers have taken drugs that affect vitamin K metabo-
lism.3~-4~For example, therapeutic doses of warfarin taken
during pregnancy may result in severe early HDN35;
warfarin is also a teratogen, and for these reasons is
contraindicated in pregnancy.
Maternal anticonvulsants have also been linked to early
HDN. 36-39Most cases have involved barbiturates, pheny-
toin, or both. These drugs may cause an increased induc-
tion of microsomal enzymes in the fetal liver.4~ These
hepatic enzymes may increase the rate of oxidative degra-
dation of the vitamin and produce a deficiency state. A
prolongation of the PT has also been observed in adults
given phenytoin.42 However, the possibility that the under-
lying disease of the mother or perinatal complications,
rather than medications, may contribute to the hemorrhag-
ic tendency in these babies has not been excluded.
The extent of bleeding varies from skin bruising36 or
umbilical bleeding3637to widespread and fatal intracranial,
intrathoracic, intra-abdominal, and gastrointestinal hem-
orrhage.38. 39At present, the exact risk for this complication
in infants born to epileptic mothers is unknown. Depressed
values of vitamin K~tependent factors have been noted in
seven of 16 consecutive infants; two of these had clinical
bleeding.36 In a survey of 1 11 such pregnancies, 20 in-
fants had prothrombin values <20% within 12 hours of
The Journal of Pediatrics
March 1985
birth? 9 Of these, eight had clinical hemorrhage, which re-
sulted in three deaths and one infant with neurologic
sequelae.
Infants born to women taking rifampin and isoniazid
during pregnancy may also be at risk for early H D N ? ~ In
six such pregnancies, three infants developed early HDN,
with one infant death and one maternal death caused by
hemorrhage? ~
Finally, rare cases of early HDN have occurred for
which no explanation is readily available. Large cephalo-
hematomas 33and scalp bleeding from the site of fetal blood
sampling34 have occurred after normal pregnancies and
have been attributed to vitamin K deficiency. We have
recently observed an idiopathic case in which a massive
intracranial hemorrhage was present at birth in a term
infant with severe vitamin K deficiency. That infant now
has severe neurologic sequelae.
Classic hemorrhagic disease of the newborn. Classic
HDN typically occurs at 2 to 5 days of a g e . 37'43'44 Affected
infants are normal at birth but subsequently develop
generalized ecchymoses or gastrointestinal bleeding. Nasal
bleeding or bleeding following circumcision may also be
the initial manifestation. Intracranial hemorrhage is less
common at this age, but sudden intracranial hemorrhage
or exsanguinating umbilical hemorrhage may occur at 2 to
3 weeks of age if the vitamin K deficiency remains
undetected. Most cases of vitamin K deficiency occurring
after the first day of life are idiopathic, although some
occur in infants born to mothers taking drugs that affect
vitamin K metabolism?7
Estimates of the incidence of classic HDN prior to the
initiation of routine vitamin K prophylaxis vary widely. It
may be as high as 1.7% in full-term infants.44 Other
estimates are somewhat lower, ranging from one in 200 to
400 term infants. The impression that classic HDN occurs
more frequently in families of low socioeconomic status is
difficult to substantiate.
Breast-feeding plays an important role in the pathogen-
esis of classic HDN. Breast milk is relatively deficient in
vitamin K. Newborn infants fed breast milk have a
significantly prolonged PT compared with those fed cow
milk or those given vitamin K at birth? ~ The incidence of
moderate to severe bleeding among breast-fed infants who
do not receive vitamin K is 15 to 20 times greater than in
infants who receive cow milk, vitamin K, or both? 4
Although some controversy regarding the need for
vitamin K prophylaxis in all newborn infants remains,
there is little doubt that classic HDN is virtually nonexis-
tent in infants given a parenteral dose of,vitamin K at
birth.
Late hemorrhagic disease. During the past 10 years it
has become apparent that vitamin K-deficiency hemor-
Volume 106
Number 3
rhage is an important cause of morbidity and mortality in
infants older than 1 month. 37'45"65 Most of these infants
have acute intracranial hemorrhage, which can be intra-
cerebral, intracerebellar, subarachnoid, subdural, or epi-
dural, as the initial feature. Often the first signs of illness
are severe central nervous system dysfunction with vascu-
lar collapse. Although excessive bleeding from puncture
sites is usually noted at presentation, there is often no
history of bleeding or bruising prior to the onset of the
illness. Many of these infants die, and those who survive
frequently have severe neurologic sequelae. In one series,
vitamin K deficiency was the leading cause of intracranial
hemorrhage in infants after the first week of life.49
The second most common initial feature of late hemor-
rhagic disease is widespread deep ecchymoses, or "nodular
purpura. ''47,55 Other less common initial manifestations
include gastrointestinal47 or mucous membrane bleeding,
as well as excessive hemorrhage following a surgical
procedure45 or intramuscular injection.
Vitamin K-deficiency hemorrhage in older infants can
be idiopathic or may occur as a secondary manifestation of
an underlying disorder? ~ Idiopathic cases typically
occur between 1 and 3 months of age, whereas secondary
cases may be seen any time during the first year. The
secondary causes of vitamin K deficiency in infants older
than 1 month are listed in Table III.
In 1946, seven infants with chronic diarrhea were
described in whom significant hypoprothrombinemia was
corrected by the administration of vitamin K. 56 Five of
these infants bled. The authors speculated on three possible
causes of vitamin K deficiency in these patients: insuffi-
cient dietary intake, decreased intestinal absorption, and
decreased vitamin K synthesis by intestinal flora secondary
to the diarrhea or to the use of antibiotics. Many additional
cases of vitamin K-deficiency hemorrhage in infants with
diarrhea have been reportedr s,ss.57 Although the role of
antibiotics has not been discounted, late hemorrhagic
disease occurs in infants with diarrhea who have not
received antimicrobial therapy? 5,57 Vitamin K deficiency
occurs in thriving infants after relatively brief bouts of
diarrhea, particularly if they have been breast-fed. 45,57
These reports have led the American Academy of Pe-
diatrics66 to recommend that "breast-fed infants who
develop diarrhea of longer than several days duration
should be given one intramuscular injection of vitamin
K .... '~
Cystic fibrosis is a well-known cause of vitamin K
deficiency at any age. In one survey, 58% of patients with
cystic fibrosis had laboratory evidence of vitamin K
deficiency.67 In infancy, vitamin K-deficiency hemorrhage
may be the first sign of cystic fibrosis58-6~ the initial
manifestation may be bruising, hematemesis, or intracra-
Vitamin K in infancy 35 5
nial hemorrhage. One child with cystic fibrosis was
referred for evaluation of possible nonaccidental trauma. 6~
In two well documented cases the correct diagnosis was
discovered only at autopsy in infants who died of sudden
intracranial hemorrhage. 5~'59
Late hemorrhagic disease in infancy has been the initial
manifestation of biliary atresia s~ s~ and arantitrypsin deft-
ciency?2.~ In some of these infants the persistence of
jaundice after the first 2 weeks of life had been attributed
to breast-feeding; it was only after a hemorrhage that
conjugated hyperbilirubinemia was detected and the cor-
rect diagnosis suspected. Bleeding secondary to vitamin K
deficiency has also been the initial feature of neonatal
hepatitis62 and abetalipoproteinemia.6~ In older infants,
celiac disease may be complicated by vitamin K deficien-
cy. 64 Because many of these putative secondary cases have
occurred in young infants who did not receive vitamin K at
birth and who were breast-fed, factors other than malab-
sorption may be significant in the cause of vitamin K
deficiency in these infants.
Finally, the possibility of chronic warfarin poisoning
must be considered. Warfarin compounds can be absorbed
transcutaneously as well as from the gastrointestinal tract.
Recently, 741 cases of vitamin K-deficiency hemorrhage,
including 177 deaths, were traced to the use of warfarin-
contaminated talcum powder. 65 On the other hand, warfa-
rin administration to the breast-feeding mother does not
appear to affect her infant adversely.6~
The importance of these secondary causes of vitamin K
deficiency is twofold. First, the possibility of a serious
underlying disease must be considered. Certainly the
history and physical examination should exclude poor
weight gain, abnormal stools, prolonged jaundice, or war-
farin exposure. In addition, a sweat chloride test, fraction-
ated bilirubin determination, and review of the peripheral
blood smear for the presence of acanthocytes is indicated.
An autopsy should be performed in fatal cases, particularly
in light of the implications in regard to genetic counseling.
Second, physicians must be aware of the need for vitamin
K supplementation in patients with diarrhea or fat malab-
sorption. The Committee on Nutrition of the American
Academy of Pediatrics has recommended a daily dietary
supplement of 50 to 100 tzg or a monthly injection of 1 mg
of vitamin K in these infants.~6
In many cases of late hemorrhagic disease in infants no
underlying cause has been identified?7,45-s4More than 200
of these idiopathic cases have been reported during the
past 10 years.alone; acute intracranial hemorrhage was the
most frequent iltitial manifestation. The overwhelming
majority of these infants have been breast-fed, often
exclusively. Preliminary data suggest that the vitamin K
content of human milk may be very low in these infants.69
356 Lane and Hathaway
In one report, the vitamin K content of the milk of nine
mothers whose infants had vitamin K-deficiency bleeding
was 4.17 _+ 2.42 gg/L, compared with 8.87 _+ 3.67 #g/L
in 13 controls.6~ Antibiotic use may play a role in the
pathogenesis of the deficiency in some instances,46'47.53but
well-documented cases occur without any history of antibi-
otic administration.48 Idiopathic, late hemorrhagic dis-
ease occurs rarely in infants who receive vitamin K at
birth? ~
TREATMENT OF VITAMIN K DEFICIENCY
In infants, 1 mg of vitamin K is adequate to correct even
a severe deficiency state. The rapidity with which this
correction occurs is explained by the fact that precursor,
noncarboxylated procoagulants are readily carboxylated in
the presence of vitamin K. The coagulopathy is usually
corrected within a few hours after a parenterally adminis-
tered dose of the vitamin. When the deficiency is severe, it
may be prudent to administer the vitamin intravenously,
because intramuscular administration may be associated
with hematoma formation as well as a slightly longer time
to correction. For life-threatening hemorrhage, parenteral-
ly administered vitamin K should be followed with 10 to 20
ml/kg of fresh-frozen plasma, which will immediately
boost the levels of the vitamin K-dependent factors.
Three points deserve emphasis. Vitamin K~leficiency
hemorrhage may be complicated by evidence of dissemi-
nated intravascular coagulation,7~particularly if intracra-
nial hemorrhage or hypovolemia has occurred. Hence, the
parenteral administration of vitamin K should never be
withheld just because of elevated values of fibrin degrada-
tion products or evidence of platelet or fibrinogen con-
sumption. When in doubt, give vitamin K. Second, the
abnormal, noncarboxylated prothrombin can be detected
in the plasma for hours after the correction of vitamin K
deficiency; the deficiency can be confirmed after treatment
has been initiated?3 Therefore, vitamin K administration
should not be delayed in a bleeding infant because of
difficulty in obtaining adequate samples for coagulation
studies. Finally, the use of vitamin K~ in infancy is safe.
Although hemolytic anemia and kernicterus have been
reported in neonates given vitamin K3 (Synkayvite) paren-
terally, such administration of vitamin K~ oxide in doses as
high as 10 to 25 mg has not been associated with toxic
symptoms or hyperbilirubinemia.9
CURRENT CONTROVERSIES
Do aH newborn infants require prophylaxis? Recent
medical literature contains numerous challenges to the
recommendation that all newborn infants receive prophy-
lactic vitamin K at birth? 9'7~-7~Many of these objections
The Journal of Pediatrics
March 1985
have been based on the failure of some investigators to find
evidence of vitamin K deficiency in small series of infants.
van Doorm et al. 7~ failed to find PIVKA in 43 consecutive
cord blood samples. Malia et al. ~9 studied 24 cord blood
samples and found no evidence of vitamin K deficiency
with several different assays. G6bel et al. 72 studied 154
healthy infants at 72 to 94 hours of age who had not
received vitamin K at birth and did not find a depressed
prothrombin level in any infant who received a feeding
during the first 24 hours of age, although none was
exclusively breast-fed. Finally, Mori et al. 73 found no
beneficial effect of vitamin K on the levels of the vitamin
K-dependent factors in 31 premature infants.
Supporters of vitamin K prophylaxis have countered
with important objections.74"77Vitamin K prophylaxis was
initiated to prevent disease that, in most series, occurred in
one in 200 to 400 neonates; conclusions based on small
samples are hazardous. AbaUi74 has suggested that the
failure to find PIVKA in the cord blood does not exclude
the possibility that vitamin K deficiency will develop 48 to
72 hours later. Also, the beneficial effect of early cow milk
or formula feedings in the prevention of classic HDN is
solidly established. The failure to find evidence of vitamin
K deficiency in infants who have received such feedings
should not be cited as reason to cease prophylaxis in other
infants.
Recent data suggest that some newborn infants are, in
fact, vitamin K deficient at birth. In one study, 15 of 25
term neonates had evidence of noncarboxylated prothrom-
bin in the plasma. ~8 Corrigan and Kryc ~6found depressed
factor II coagulant/antigen ratios in seven of 40 term
infants but suggested that this did not occur in the absence
of perinatal complications. However, Yoshioka et al? ~
found decreased ratios in six of eight normal term infants
at 3 days of age. These infants did not have perinatal
complications, did not receive vitamin K at birth, and were
breast-fed. In a preliminary report, Blanchard et al. TM
applied their sensitive immunoassay for abnormal pro-
thrombin to 99 pregnant women and the cord blood of their
infants. They found evidence of vitamin K deficiency in
33% of the mothers and 75% of the infants. In another
study of 50 mother-infant pairs, PIVKA was detected in
seven mothers and 13 infants.79
Perhaps immaturity of the Carboxylase system in the
liver, rather than vitamin K deficiency, accounts for the
abnormal prothrombin levels in some neonates. Although
this hypothesis deserves further investigation, at present
we are aware of no data to support it. Vitamin K deficiency
in maternal plasma and the poor placental transport of
vitamin K z~ argue for a true vitamin deficiency. In addi-
tion, Aballi76 has summarized a large body of convincing
Volume 106
Number 3
data showing that the low levels of vitamin K-dependent
factors present in many newborn infants respond dramati-
cally to vitamin K administration.
A 1978 editorial in The Lancet 8~ suggested that prophy-
laxis might be safely withheld from term infants who have
no perinatal complications, are not breast-fed, have no
obvious bleeding, do not receive antibiotics or require
surgery, and will not be discharged from the nursery early.
We agree that breast-fed infants and those with perinatal
complications are a t high risk for vitamin K deficency. We
do not agree that adequate data are available to support
the exclusion of any neonate from routine prophylaxis.
Recent experience in England, where cessation of routine
prophylaxis has led to a resurgence of vitamin K~cleficien-
cY bleeding, supports the routine administration of vitamin
K to all newborn infants? 7
Is the intramuscular injection necessary for prophylaxis?
The possibility that effective prophylaxis could be admin-
istered orally has been suggested. 8 Theoretic objections
include the possibility of variable absorption, the possibili-
ty that regurgitated doses might go Unnoticed and not be
repeated, and the potential risk of lipid aspiration. In
addition, a somewhat later onset of effect might increase
the risk of early hemorrhage in infants already vitamin K
deficient at birth. In a preliminary report, the oral admin-
istration of 2 mg vitamin K~ to neonates effected a marked
increase in the plasma level of vitamin K? 2 Dunn has cited
experience in which 1 mg vitamin K1 was added to the first
feeding of 31,000 newborn infants, 77 with only one case of
hemorrhage in an infant of a mother taking anticonvulsant
drugs. Whether any of these infants was exclusively
breast-fed i s not mentioned. In a study from Japan,
vitamin K was administered orally to newborn infants with
hypoprothrombinemia, s~ The response to 2 mg vitamin K~
was not different from that to a placebo; however, vitamin
K2 appeared to be somewhat more effective. Further study
is needed to document the efficacy, safety, and optimal
dosage of this approach.
The early studies of Dam et al. 6 indicated that vitamin K
administration to the mother prior to delivery was less
effective in preventing hypoprothrombinemia in the infant
unless given within hours of delivery. The transplacental
passage of vitamin K is inefficient,2L and the possibility of
effective prophylaxis by this method needs further investi-
gation.
What is the optimal management of infants at risk for
early HDN? Early H D N can occur quickly after delivery,
and death has occurred despite the prompt intramuscular
administration of vitamin K to the infant) ~ In particular,
there is concern when anticonvulsant drugs are required by
the mother during pregnancy. In 1973 Seip 8z suggested the
Vitamin K in infancy 357
daily administration of 10 mg vitamin K orally to the
mother during the last 2 months of pregnancy. Bleyer and
Skinner, 3g in their review of 21 cases of hemorrhagic
disease, recommended that women taking anticonvulsant
agents during pregnancy avoid drugs such as salicylates
and that they be given vitamin K prior to delivery. These
authors also suggested that Caesarean section be consid-
ered if a difficult or traumatic delivery is anticipated.
Finally, they recommended immediate cord blood clotting
studies as well as the intravenous administration of vitamin
K to the infant in the delivery room. Preliminary data
suggest that the oral administration of vitamin K to the
mother (20 mg/daY for 2 weeks) prior to delivery may be
sufficient to prevent a deficiency in the infant. 39 Although
these suggestions are empirically sound, we are aware of no
data documenting their efficacy. We have combined t h e
immediate intramuscular administration of vitamin K to
the infant with close observation. Further studies are
needed befor e any firm guidelines for the perinatal man-
agement of these pregnancies can be established.
Should normal infants who are exclusively breast-fed
receive vitamin K supplementation after the immediate
newbor~ period? This question has been raised by the large
number of cases of idiopathic late hemorrhagic disease in
breast-fed infants. Although the majority o f these infants
did not receive vitamin K at birth, rare cases have occurred
at 4 to 8 weeks of age in infants who did receive
prophylaxis? LThe incidence of vitamin K deficiency in 4-
to 12-week-old breast-fed infants who have received vita-
min K at birth is unknown. Jimenez et al. 83 studied infants
at l month of age who had received vitamin K prophylaxis
and were exclusively breast-fed. One of 21 infants had a
significantly decreased prothrombin level. Vitamin K
administration to the lactating mother may significantly
increase the vitamin K content of her milk54 Further study
of vitamin K supplementation to lactating mothers or to
breast-fed infants is needed.
CONCLUSIONS
Considerable recent progress has been made in our
understanding of the role of vitamin K in infancy. TheSe
recent advances strongly support the continuation of rou-
tine newborn prophylaxis. In the United States, wide-
spread acceptance of this important public health measure
has been associated with a dramatic decrease in the
incidence of classic hemorrhagic disease of the newborn
infant. Nevertheless, some infants are at risk for severe
vitamin K-defxoiency hemorrhage at the time of birth. The
problem of vitamin K~leficiency intracranial hemorrhage
in older infants has been recognized worldwide as an
important cause of morbidity and mortality. Physicians
358 Lane and Hathaway
who care for children m u s t r e m a i n alert to the possibility
of vitamin K deficiency in older infants and search
carefully for underlying causes when it occurs. W e hope
t h a t ongoing research will provide answers to the clinically
relevant questions t h a t remain.
We thank Ms. Norma Kure for assistance in preparation of the
manuscript.
REFERENCES*
1. Townsend CW: The haemorrhagic disease of the newborn.
Arch Pediatr 11:559, 1894.
2. Dam H: Cholesterinstoffwechsel in hiihnereiern und h(ihnch-
en. Biochem Zeitschr 215:475, 1929.
3. Brinkhous KM, Smith HP, Warner ED: Plasma prothrombin
level in normal infancy and in hemorrhagic disease of the
newborn. Am J Med Sci 193:475, 1937.
4. Waddell WW, Guerry D: The role of vitamin K in the
etiology, prevention, and treatment of hemorrhage in the
newborn infant. J PEDIATR 15:802, 1939.
5. Nygaard KK: Prophylactic and curative effect of vitamin K in
hemorrhagic disease of the newborn. Acta Obstet Gynecol
Scand 19:361, 1939.
6. Dam H, Dyggve H, Larsen H, et al: The relation of vitamin K
deficiency, to hemorrhagic disease of the newborn. Adv
Pediatr 5:129, 1952.
7. Aballi A], de Lamerens S: Coagulation changes in the
neonatal period and in early infancy. Pediatr Clin North Am
9:785, 1962.
8. Wefring KW: Hemorrhage in the newborn and vitamin K
prophylaxis. J PEDIATR 61:686, 1962.
9. Committee on Nutrition, American Academy of Pediatrics:
Vitamin K compounds and the water-soluble analogues: Use
in )herapy and prophylaxis in pediatrics. Pediatrics 28:501,
196i.
10. Uotila L, Suttie JW: Recent findings in understanding the
biological function of vitamin K. Med Biol 60:16, 1982.
11. Liebman HA, Furie BC, Furie B: Hepatic vitamin K -
dependent carboxylation of blood-clotting proteins. Hepatolo-
gy 2!488, 1982.
12. Esmon CT: Protein-C: Biochemistry, physiology, and clinical
impiications. Blood 62:1155, 1983.
13. Gallop PM, Lian JB, Hauschka PV: Carboxylated calcium-
binding proteins and vitamin K. N Engl J Med 302:1460,
1980.
14. Stenflo J, Fernlund P, Egan W, et al: Vitamin K-dependent
modifications of glutamic acid residues in prothrombin. Proc
Natl Acad Sci USA 71:2730, 1974.
15. Hathaway WE, Bonnar J: Perinatal coagulation. New York,
1978, Grune & Stratton, p 53.
16. Corrigan J J, Kryc J J: Factor II (prothrombin) levels in cord
blood: Correlation of coagulant activity with immunoreactive
protein. J PEDIATR 97:979, 1980.
17. Meguro M, Yamada K: A simple and rapid test for PIVKA-II
in plasma. Thromb Res 25:109, 1982.
18. Muntean W, Petek W, Rosanelli K, et al: Immunologic
*Because of space limitations, the original list of references has been
significantly shortened. The authors will provide a full list of references to
interested readers on request.
The Journal of Pediatrics
March 1985
studies of prothrombin in newborns. Pediatr Res 13:1262,
1979.
19. Malia RG, Preston FE, Mitchell VE: Evidence against
vitamin K deficiency in normal neonates. Thromb Haemost
44:159, 1980.
20. Blanchard RA, Furie BC, .!orgensen M, et al: Acquired
vitamin K~zlependent carboxylation deficiency in liver dis-
ease. N Engl J Med 305:242, 1981.
21. Shearer M J, Rahim S, Barkhan P, et al: Plasma vitamin K1 in
mothers and their newborn babies. Lancet 2:460, 1982.
22. Sanford HN, Gesteyer TH, Wyat L: The substances involved
in the coagulation of blood of the newborn. Am J Dis Child
43:58, 1932.
23. Corrigan J J: The vitamin K-dependent proteins. Adv Pediatr
28:57, 1981.
24. Haroon Y, Shearer M J, Rahim S, et al: The content of
phylloquinone (vitamin K1) in human milk, cow's milk and
infant formula foods determined by high-performance liquid
chromatography. J Nutr 1i2:1105, 1982.
25. Hollander D: Intestinal absorption of vitamins A, E, D, and
K. J Lab Clin Med 97:449, 1981.
26. Bentley R, Meganathan R: Biosynthesis of vitamin K (meua-
quinone) in bacteria. Microbiol Rev 46:241, 1982.
27. Aballi A J, Howard CE, Triplett RF: Absorption of vitamin K
from the colon in the newborn infant. J PEDIATR 68:305,
1966.
28. Keenan W J, Jewett T, Glueck HI: Role of feeding and
vitamin K in hypoprothrombinemia of the newborn. Am J Dis
Child 121:271, 1971.
29. Bjornsson TD, Meffin P J, Swezey SE, et al: Disposition and
turnover of vitamin K~ in man. In Suttie JW, editor: Vitamin
K metabolism and vitamin K-dependent proteins. Baltimore,
1980, University Park Press, p 328.
30. Frick PG, Riedler G, Br6gli H: Dose response and minimal
daily requirement for vitamin K in man. J Appl Physiol
23:387, 1967.
31. Shearer M J, Barkhan PK Webster G R : Absorption and
excretion of an oral dose of tritiated vitamin K~ in man. Br J
Haematol 18:297, 1970.
32. Sann L, Leclercq M, Bourgeois J, et al: Pharmacokinetics of
vitamin Ka in newborn infants [abstract]. Pediatr Res
17:155A, 1983.
33. Leonard S, Anthony B: Giant cephalhematoma of newborn
with hemorrhagic disease and hyperbilirubinemia. Am J Dis
Child 101:170, 1961.
34. Hull MG, Wilson JA: Massive scalp haemorrhage after fetal
blood sampling due to haemorrhagic disease. Br Med J 4:321,
1972.
35. Stevenson RE, Burton OM, Ferlauto G J, et al: Hazards of
oral anticoagulants during pregnancy. JAMA 243:1549,
1980.
36. Mountain KR, Hirsh J, Gallus AS: Neonatal coagulation
defect due to anticonvulsant drug treatment in pregnancy.
Lancet 1:265, 1970.
37. McNinch AW, Orme RL'E, Tripp JH: Haernorrhagic disease
of the newborn returns. Lancet 1:1089, 1983.
38. Bleyer WA, Skinner AL: Fatal neonatal hemorrhage after
maternal anticonvulsant therapy. JAMA 235:626, 1976.
39. Deblay MF, Vert P, Andre M, et al: Transplacental vitamin
K prevents haemorrhagic disease of the infant of epileptic
mother [letter]. Lancet 1:1247, 1982.
40. Eggermont E, Logghe N, van de Casseye W, et al: Haemor-
Volume 106
Number 3
rhagic disease of the newborn in the offspring of rifampin and
isoniazid treated mothers. Acta Paediatr Belg 29:87, 1976.
41. Keit h DA, Gallop PM: Phenytoin, hemorrhage, skeletal
defects and vitamin K in the newborn. Med Hypotheses
5:1347, 1979.
42. Andreasen PB, Lyngbye J, Trolle E: Abnormalities in liver
function tests during long-term diphenylhydantoin therapy in
epileptic outpatients. Acta Med Scand 194:261, 1973.
43. Aballi A J: Hemorrhagic disease of the newborn. Pediatr Ann
3(2):35, 1974.
44. Sutherland JM, Glueck HI, Gleser G: Hemorrhagic disease
of the newborn: Breast feeding as a necessary factor in the
pathogenesis. Am J Dis Child 113:524, 1967.
45. Lovric VA, Jones RF: The haemorrhagic syndrome of early
childhood. Aust Ann Med 16:173, 1967. \ 55.
46. Visndhiphan P, Bhanchet P, Lakanapichanchat C, et al:
Intracranial hemorrhage in infants due to acquired prothrom-
bin complex deficiency. J Neurosurg 41:14, 1974.
47. Bhanchet P, Tuchinda S, Hathirat P, et al: A bleeding
syndrome in infants due to acquired prothrombin complex
deficiency: A survey of 93 affected infants. Clin Pediatr
i6:992 , 1977.
48. Cooper NA, Lynch MA: Delayed haemorrhagic disease of the
newborn with extradural haematoma. Br Med J 1:164, 1979.
49. Matsusaka T, Katayama K, Tsuji Y, et al: Intracranial
hemorrhage due to vitamin K deficiency in infants: Its
importance as a cause of intracranial hemorrhage in infants.
No To Sbinkei 33:989, 1981.
50. Fujimura Y, Mimura Y, Kinoshita S, et al: Studies on
vitamin K-dependent factor deficiency during early child-
hood with special reference to prothrombin activity and
antigen level. Haemostasis 11:90, 1982.
51. Verity CM, Carswell F, Scott GL: Vitamin K deficiency
causing infantile intracranial haemorrhage after t h e neonatal
period [letter]. Lancet 1:1439, 1983.
52. O'Connor ME, Livingston DS, Hannah J, et al: Vitamin K
deficiency and breast-feeding. Am J Dis Child 137:601,
1983.
53. Lane PA, Hathaway WE, Githens JH, et al: Fatal intracra-
nial hemorrhage in a normal infant secondary to vitamin K
deficiency. Pediatrics 72:562, 1983.
54. Yoshioka K, Kinoshita S, Takamiya O, et al: Abnormal
antigens of factors 1I, VII, and IX in newborns and breast-fed
infants with vitamin K deficiency. Acta Haematol Jpn
45:860, 1982.
55. Nagi NA, A1-Dubooni HM, A1-Shirkat SAR: Nodular pur-
pulra in infancy. Postgrad Med J 58:274, 1982.
56. Rapaport S, Dodd K: Hypoprothrombinemia in infants with
diarrhea. Am J Dis Child 71:611, 1946.
57. Matoth Y: Plasma prothrombin in infantile diarrhea. Am J
Dis Child 81t:944, I950.
58. Torstenson OL, Humphrey GB, Edson JR, et al: Cystic
fibrosis presenting with severe hemorrhage due to vitamin K
malabsorption: A report of 3 cases. Pediatrics 45:857, 1970.
59. Oppenheimer EH, Schwartz A: Easy bruisability and termi-
nal coma in a "normal" 5-month-old infant. J PEDIATR
88:1049, 1976.
60. Carpentieri U, Gustavson LP, Haggard ME: Misdiagnosi s of
neglect in a child with bleeding disorder and cystic fibrosis.
South Med J 71:854, 1978.
61. Hope PL, Hall MA, Millward-Sadler GH, et al: Alpha-
Vitamin K in infancy 359
1-antitrypsin deficiency presenting as a bleeding diathesis in
the newborn. Arch Dis Child 57:68, 1982.
62. Poley JR, Humphrey GB: Bleeding disorder in an infant
associated with anicteric hepatitis. Clin Pediatr 13:1045,
1974.
63. Caballero FM, Buchanan GR: Abetalipoproteinemia present-
ing as severe vitamin K deficiency. Pediatrics 65:161, 1980.
64. Rodesch P, Cadranel S, Winckler M, et al: Hypovitaminosis
K in coeliac disease. Acta Paediatr Belg 29:123, 1976.
65. Martin-Bouyer G, Khanh NB, Linh PD, et al: Epidemic of
haemorrhagic disease in Vietnamese infants caused by warfa-
rin-contaminated talcs. Lancet 1:230, 1983.
66. Barness LA: Vitamins. In Pediatric nutrition handbook.
Evanston, I11., 1979, American Academy of Pediatrics, p
67. Corrigan J J, Taussig LM, Beckerman R, et al: Factor II
(prothrombin) coagulant activity and immunoreactive pro-
tein: Detection of vitamin K deficiency and liver disease in
patients with cystic fibrosis. J PEDIATR 99:254, 1981.
68. McKenna R, Cole ER, Vasan U: Is warfarin sodium contra-
indicated in the lactating mother? J PED1ATR 103:325,
1983.
69. Isarangkura PB, Mahadandana C, Panstienkul B, et al:
Vitamin K level in maternal breast milk of infants with
acquired prothrombin complex deficiency syndrome. South-
east Asian J Trop Med Public Health 14:275, 1983.
70. Mangnrten HH, Honig GR: Hemorrhagic disease due to
vitamin K deficiency in a premature infant: A syndrome
which may resemble disseminated intravascular coagulation
(DIC). Clin Pediatr 12:372, 1973.
71. van Doorm JM, Muller AD, Hemker HC: Heparin-like
inhibitor, not vitamin-K deficiency, in the newborn [letter].
Lancet 1:852, 1977.
72. G/3bel U, Sonnenschein-Kosenow S, Petrich C, et al: Vitamin-
K deficiency in the newborn [letter]. Lancet 2:187, 1977.
73. Mort PG, Bisogni C, Odino S, et al: Vitamin-K deficiency in
the newborn [letter]. Lancet 2:188, 1977.
74. Aballi A J: Vitamin K deficiency in the newborn [letter].
Lancet 2:559, 1977.
75. Edson JR: Vitamin-K deficiency in the newborn [letter].
Lancet 2:187, 1977.
76. Aballi A J: Vitamin K and the newborn [letter]. Lancet
1:1358, 1978.
77. Dunn PM: Vitamin K1 for all newborn babies [letter]. Lancet
2:770, 1982.
78. Blanchard RA, Furie BC, Peck C, et al: Subclinical vitamin K
deficiency in newborns and their mothers. [abstract] Blood
62(Suppl 1):272a, 1983.
79. Bloch CA, Rothberg AD, Bradlow BA: Mother-infant pro-
thrombin precursor status at birth. J Pediatr Gastroenterol
Nutr 3:101, 1984.
80. Vitamin K a n d the newborn [editorial]. Lancet 1:755,
1978.
81. Suzuki S, Maki M, Shirakawa K, et al: The effects of vitamin
K2 syrup on neonatal hypoprothrombinemia: Double blind
study. Acta Haematol Jpn 45:876, 1982.
82. Seip M: Effects of antiepileptic drugs in pregnancy on the
fetus and newbo[n infant. Ann Clin Res 5:205, 1973.
83. Jimenez R, Navarette M, Jimenez E, et al: Vitamin K-
dependent clotting factors in norma[~'breast-fed infants. J
PEDIATR 100:424, 1982.