484 EDITORIALS
GASTROENTEROLOGY Vol. 121, No. 2
is more closely associated with the origin and socioeco-
nomic status of the mother over the father as evidenced
by an odds ratio of 71.
upper endoscopy. Antral specimens were obtained and
the presence of H. pylori was determined. One of the 12
infants was found to be infected with H. pylori. Despite
the fact that only 1 of 20 children was infected with H.
pylori, 19 of the 20 mothers were positive by the CLOtest
and histopathological assessment, including the mother
of the infant who tested positively for H. pylori. All
children in the study were breast-feeding. Nine of the
children were receiving complimentary formula in addi-
tion to breast-feedings, including the infant infected
with H. pylori. Despite close contact between H. pylori–
positive mothers and their children, an extremely low
prevalence of H. pylori infection was found in the chil-
dren studied. If the Tindberg hypothesis is correct, this
study suggests that transmission from mother to child
would have to occur after the age of weaning.
Further analysis revealed that the prevalence of H.
pylori infection among infected classmates was in no way
determinative of the likelihood of being infected. Chil-
dren born in Sweden to parents also born in low preva-
lence areas were examined. Those attending school with
a seroprevalence of 30% were no more likely to be
infected than children attending school with a seropreva-
lence of Ͻ10% (odds ratio, 1.1). These findings are
consistent with the many previous studies documenting
the low level of H. pylori acquisition in children over 5
years old, regardless of environment and family status.
The present report and previous studies all support the
widely held belief that infection with H. pylori most
likely occurs in children less than 5 years old.
Goodman et al. evaluated just such an age group by
studying the effects of the family composition on the H.
pylori prevalence in children aged 2–9 years in a rural area
of the Columbian Andes. They showed a clear gradient of
effect seen in birth order.11 When the investigators ex-
amined the effects of birth spacing, they showed that
children born within 4 years of an older sibling were 4
times more likely to be infected with H. pylori. The odds
of having H. pylori infection seems to increase almost
steadily with the total number of 2–9-year-old siblings
in the home. This effect seems to depend entirely on the
number of older siblings in the age range. The investi-
gators concluded that birth order, birth spacing, and the
infectious status of siblings influenced the odds of ac-
quiring H. pylori infection independently of the number
of children in the home and suggests that the organism
is most readily transmitted among the siblings who are
close in age and more frequently from older to the
younger ones. This is in contrast to the results of Tind-
berg et al., which show that the intrafamilial transmis-
sion from a parent, and more specifically from mother to
child, is the most significant risk factor for the acquisi-
tion of H. pylori infection.
The present study by Tindberg et al. provides strong
evidence that the risk of H. pylori acquisition by school-
age children from extraneous child-to-child contact is
low and provides compelling data to support the notion
that transmission primarily occurs from mother to infant
child. Future studies in this area will require an in-depth
analysis of the parent-to-child and sibling-to-sibling
transmission in children under the age of 5 years old.
Such studies would go a long way in identifying the
primary means of infection. Until the means of acquisi-
tion and transmission for H. pylori infection are better
defined, we should be cognizant of the potential exposure
Perhaps the most intriguing piece of information ob-
tained in this study, and one that supports their conclu-
sion regarding the importance of intrafamilial association
over extraneous child-to-child contact for H. pylori trans-
mission, comes from an analysis of the day care history of
these children. Almost 90% of the children assessed in
this study had attended day care, yet no increased risk for
H. pylori infection was observed compared with children
that had received their care exclusively at home. These
results are informative because day care is often provided
to infants and children still believed to be within the
window of vulnerability to H. pylori acquisition. More
information as to the age during which day care was
received, the duration, and the seroprevalence of the
other children would be helpful in drawing a firm con-
clusion. But the data presented here do indicate that
acquisition most likely occurs within the family.
Unfortunately, Tindberg et al.9
studied a cohort of
10–12-year-old children and did not specifically evaluate
the acquisition of H. pylori in the first years of life.
Helicobacter pylori, like most enteric infections, is acquired
during childhood. There have been a few recent studies
that address the acquisition of H. pylori during the first
few years of life. Guelrud et al.10 assessed the prevalence
of H. pylori in a group of neonates and young infants from
a lower socioeconomic background. Antral biopsy spec-
imens were obtained from 20 consecutive neonates and
infants under the age of 2 months who were undergoing
endoscopic retrograde cholangiopancreatography (ERCP)
for an evaluation of neonatal cholestasis. Unlike the
study by Tindberg et al. in which the actual H. pylori
status of the mothers was not determined, in Guelrud et
al.’s study, on the same day that ERCP examination of
the child was conducted, the mother also underwent