J. Hyg., Camb.
(1974), 72, 67 67
Printed in Great Britain
The origin of 0 serotypes of Escherichia coli in babies
after normal delivery
BY K. A. BETTELHEIM, ALWENA BREADON, MARY C. FAIERS,
SHEILA M. O'FARRELL AND R. A. SHOOTER
Department of Medical Microbiology, St Bartholomew's Hospital,
West Smithfield, London ECIA 7BE
(Received 25 May 1973)
SUMMARY
A total of 2525 strains of Escherichia coli were isolated from the faeces of 33
mothers, the faeces of their babies and the mucus extracted from the babies'
mouths after delivery. Of these strains 1832 could be 0-serotyped with 150 0 anti-
sera. E. coli were isolated from 28 babies and the same serotypes as were found in
their mothers were found in 22 of them. E. coli was isolated from only 14 of the
mucus specimens but in 12 of these at least one of the serotypes present was
subsequently found in the babies' stools.
INTRODUCTION
Babies have Escherichia coli in their stools within a few days of birth. Some of
these organisms are thought to come from the mother's bowel at the time of
delivery but their appearance in babies delivered by Caesarean section and the
spread of entero-pathogenic strains of E. coli make it clear that intestinal organisms
of this group can be acquired from other sources.
MATERIALS AND METHODS
Full clinical details of mothers and babies were recorded. An attempt was made
to obtain stool specimens from mothers before delivery and from the baby on each
day after delivery. Mucus was also sucked from each baby's mouth immediately
after birth. Stools and mucus were sown directly on MacConkey's medium. A second
culture on a MacConkey plate was made from the mucus after it had been incubated
overnight with broth.
All plates were incubated overnight and from each at least ten colonies of E. coli
were isolated and were 0-serotyped using 150 antisera by the method described
by Bettelheim & Taylor (1969).
RESULTS
Thirty-three babies and their mothers were studied. All the babies were delivered
normally and Table 1 shows the origin of the 2525 strains of E. coli that were
isolated from them and their mothers.
5-2
68 K. A. BETTELHEIM AND OTHERS
Table 1. Distribution of strains of Escherichia coli
Number of strains from
Mothers' Babies' Babies'
Strains stools stools mucus All three
Typable with 150 0 sera 622 987 223 1832
Not typable 154 151 2 307
Rough 145 215 26 386
Total 921 1353 251 2525
The babies were observed for an average of 7 days, and in this time 28 were
found to have one or more strains of E. coli in the stool. In 22 the same serotype
had been found in the mother's stool either before delivery (19 cases) or after
delivery (21 cases). For the present study rough or non-typable strains were
considered similar if they were found in specimens from one mother and her baby.
Mucus was examined from 30 of the 33 babies and 14 of these samples contained
E. coli. In 12 of these at least one of the serotypes present was found subsequently
in the babies' stools, and in 11 of these 12 at least one of the serotypes was also
found in the mother's stools (Table 2).
More serotypes were identified in mothers' stools than in babies' stools, 119 sero-
types being identified from 32 mothers and 60 from the 28 babies whose stools
were positive. The number of serotypes present in the babies appeared to be
related to the presence or absence of E. coli in the baby's mucus. For the 14 babies
with E. coli present in the mucus, 40 serotypes were identified in the stools com-
pared with 19 serotypes for the 16 babies with no E. coli in the mucus. Only one
of the five babies from whose stools no E. coli were isolated had the organism
present in the mucus.
All the babies were given supplementary bottles and eight were breast fed, but
this did not appear to be related. Of more interest was the finding that in the
11 babies from whom maternal serotypes were not isolated, the mean time from
rupture of the membranes to birth was 157 min., as compared with 284 min. for
the 22 babies from whom maternal serotypes were isolated.
DISCUSSION
Since Escherich (1885) first observed that, although sterile at birth, babies'
stools soon contained bacteria, including E. coli, there have been many studies of
the babies' faecal flora, and in recent years particular attention has been paid to
the presence of entero-pathogenic strains.
The use of an almost complete set of E. coli 0-sera has enabled us to identify
almost certainly the same serotype in mother and baby in 22 of the 28 babies from
whose stools we isolated E. coli. Our results therefore support and extend those of
others (Nejedla, Srajbr & Lanc, 1967; Gareau et al. 1959; Rosner, 1966; Ironside,
Brennand, Mandal & Heyworth, 1971 ) who have claimed that at least a proportion
of the baby's strains come from the mother, probably at the time of delivery, and
additional support for this view is given by the frequency with which we found
O serotypes of E. coil in babies 69
Table 2. O-serotypes in mothers and babies
Babies'
, ~~~~~A --"
Pt. no. Mothers' stools Mucus Stools
2 O 3, 0 7, 0 8, 0 19, 0 3, 0 78, 0 81 0 3, 0 81
0 78, 0 81, NT
17 O 18, 0 71, 0 81, R 0 81, 0 147, R 71, 0 81
0
21 O 10, 0 18, 0 58, R 0 10, 0 18, 0 58, R 18, 0 58, R
0
27 O 27, 0 40, 0 108, R, 0 19, 0 27, 0 41, 0 108, 0 27, 0 79, 0 108,
NT 0 116, 0 129, R 0 116, R, NT
35 O 11, R 0 11, R 0 1, 0 11, R
1 O 1, 0 8, 0 102, NT No specimen 0 1, R
7 0 42, 0 96, R No E. coli 0 20, 0 42, R
15 O 4, 0 12, R No E. coli R
19 0 1, R No E. coli R
26 O 1, 0 46 No E. coli 01
31 O 9, 0 16, 0 129, NT No E. coli 0 9
8 0 19, 0 82, 0 129, NT No. E coli No E. coli
20 06, R 0 6 No E. coli
28 0 22 No E. coli No E. coli
12 0 145, R No E. coli NT
16 0 6, 0 30 No E. coli 0 3
32 0 21, 0 1 No E. coli 0 3
5 0 18, 0 42, 0 78, R 0 42, 0 75, 0 78, 0 19, 0 42, 0 75, 0 78,
0 117
18 01, 0 25, 0 48, 0 71, 0 48, R 0 48, R
NT
23 0 7, 0 19, 0 46, R, NT 0 19, 0 75, R R, NT
29 0 11, 0 42, 0 79, 0 86, 0 108, R 0 42, 0 79, 0 86, 0 34,/
0 108, R, NT 0 41, R, NT
33 0 27, 0 84, 0 153, NT 0 2, 0 27, NT 0 3, 0 84, R, NT
37 0 25, R, NT 0 25 0 25, NT
6 0 7, 0 8, 0 38, 0 184, No E. coli 0 38
R
10 NT No specimen NT
13 0 5, 0 7, 0 8, 0 82, No specimen 0 7, 0 82
0 148, NT
22 0 6, 0 52, R No E. coli 0 6
30 0 46, 0 141, R, NT 0 15 0 141, NT
4 0 18, 0 26, 0 37, No E. coli No E. coli
0 75, NT
14 04, 050, 052, No E. coli No E. coli
0 147, R
3 0 3, 0 9, 0 16, 0 18, No E. coli 0 7, 0 102
0 79, 0 106, NT
24 0 9, 0 106, NT No E. coli 0 3
25 No specimen 0 10 0 10
matching of the maternal faecal serotypes with those found in the mucus from
the baby's mouth immediately after delivery.
Nine of our babies had serotypes in their stools that we did not find in the mucus
in their mouths at birth or in their mothers' stools and on six occasions serotypes
were found in the baby's mucus that we did not identify in the mother's stools.
These may represent strains acquired in some other way but present methods do
70 K. A. BETTELHEIM AND OTHERS
not permit the identification of all serotypes in a stool (Bettelheim, Faiers &
Shooter, 1972) and these strains may have been in the mother's bowel.
The study of babies delivered by Caesarean section appears to offer a chance to
investigate some of the other sources from which newborn babies acquire their
intestinal E. coli, and work on this is in progress.
We are indebted to the Wellcome Trust and to the Board of Governors of
St Bartholomew's Hospital for a grant in support of this work and we would like
to express our thanks to Miss M. Pollock and her staff in the Department of
Midwifery for their diligent help in the collection of specimens.
REFERENCES
BETTELHIEIM, K. A., FAIERS, M. & SHOOTER, R. A. (1972). Serotypes of Escherichia coli in
normal stools. Lancet ii, 1224-6.
BETTELHEIM, K. A. & TAYLOR, J. (1969). A study of Escheriehia coli isolated from chronic
urinary infection. Journal of Medical Microbiology 2, 225-36.
ESCIMRICH, T. (1885). Die Darmbakterien des Neugeborenen und Sauglings. I, II. Fortschritte
der Medizin 3, 515-22, 547-54.
GAREAU, F. C., MACKEL, D. C., BORING III, J. R., PAYNE, F. J. & HAMMETT, F. L. (1959).
The acquisition of fecal flora by infants from their mothers during birth. Journal of
Pediatrics 54, 313-18.
IRONSIDE, A. G., BRENNAND, J., MANDAL, B. K. & HEYWORTH, B. (1971). Cross-infection in
infantile gastroenteritis. Archives of Disease in Childhood 46, 815-18.
NEJEDLA, Z. 8RAJBR, E. & LANc, A. (1967). Uber die Typenbestimmung von E. coli-Staimmen
in Stiihlen von Sauglingen wahrend des ersten Lebenshalbjahres und fiber die Antikorper-
bildung gegen die festgestellten Typen. Zeitschrift fur Immunitdtsforschung, Allergie und
klinische Immunologie 133, 134-43.
ROSNER, R. (1966). Antepartum culture findings of mothers in relation to infantile diarrhoea.
American Journal of Clinical Pathology 45, 732-6.
