Secondary Logo

Journal Logo


Postnatally acquired cytomegalovirus infection via breast milk: effects on hearing and development in preterm infants


Author Information
The Pediatric Infectious Disease Journal: April 2004 - Volume 23 - Issue 4 - p 322-327
  • Free



Human milk is a carefully engineered substance with unique composition that is ideal for the growth and development of infants. 1 For preterm infants in particular, nutritional as well as immunologic advantages have been ascribed to breast milk. It has, for example, a protective effect against severe infections. 2 There is a high risk of postnatal transmission of human cytomegalovirus (CMV) to the infant via breast milk because so many seropositive mothers have virus reactivation during lactation and subsequent shedding of CMV in breast milk. 3–5 Previous studies have demonstrated that the rate of transmission by consuming CMV-positive breast milk is as high as 38% in preterm infants. 3–5

In term infants the neonatal clinical course of postnatally acquired CMV infection is usually asymptomatic or very mild, whereas in preterm infants clinical symptoms are more frequent. 6, 7 In one cohort of very low birth weight (VLBW) infants, 12% of the infants with early postnatal CMV infection developed serious neonatal problems with sepsis-like symptoms (SLS). 4

There is little information about long term consequences of early postnatally acquired CMV infection in VLBW infants. Paryani et al. 8 investigated a cohort of term and preterm infants with early postnatal CMV infection. Their data suggest that pre-term infants who were infected very early in life (first 2 months) may have an increased risk of neurodevelopmental delay, cognitive impairment and impaired neuromotor function.

The aim of this study was to compare neurodevelopmental outcome, sensorineural hearing function and anthropometric measures in preterm infants with early postnatally acquired CMV infection, transmitted via breast milk, with a group of CMV-seronegative preterm infants. This study reports findings at first follow-up of subjects who are members of the prospective study for whom both Hamprecht et al. 3 and Maschmann et al. 4 in 2001 described early postnatal transmission of cytomegalovirus via breast milk.



All subjects were born between June 1995 and June 1998 and treated at the Neonatal Intensive Care Unit at University Children’s Hospital, Tubingen, Germany. Informed consent was obtained from the parents. The population of this follow-up study consisted of 22 preterm infants born before 32 weeks of gestation or with a birth weight of <1500 g (VLBW), with early postnatally acquired CMV infection. All infants had been infected via CMV-positive breast milk. They are members of a study that aims to investigate the CMV transmission rate via breast milk and onset of infection in premature infants 3–5 with long term outcome. On the day of delivery, venous blood from the mother and cord blood from the baby were taken for CMV antibody tests (IgM, IgG). Ear canal and throat swabs were obtained from the baby immediately after birth to determine the presence of the virus by virus culture and DNA PCR (sensitivity of CMV DNA detection from plasma was 300 genome equivalences/ml or 1.5 to 10 copies/5 μl of reaction volume). Congenital CMV infection was an exclusion criterion. In all participants of the study, from Day 7 onwards breast milk was sampled from the mothers, and on the same day urine from the baby was obtained. Both specimens were examined for CMV by culture and DNA PCR. Until discharge, tests for cytomegalovirolactia and cytomegaloviruria were performed every 14 days. At the corrected age of 3 months, a urine sample from the baby was obtained and examined by PCR culture.

When CMV excretion in breast milk was detected, weekly screening of breast milk and the baby’s urine was done. This allowed determination of the onset and duration of virolactia and onset of viruria. In addition serologic tests for CMV in the mother were repeated, and specimens of the mother’s saliva and urine were examined by PCR and cultures.

Transmission of CMV via breast milk was assumed when the baby exhibited signs of infection after having consumed CMV-positive breast milk and when intranatal infection had been ruled out (negative CMV DNA in surface swab) as well as congenital infection (negative CMV IgM in cord blood and initially negative urine PCR) and horizontal infection (blood products from CMV-negative donors). A detailed description of the laboratory methods has been published previously. 3–5

The control group consisted of 22 CMV-negative infants who had the same laboratory investigations as the CMV-positive group, and there was no CMV transmission in the first 6 months of life. They were individually matched for gestational age, birth weight, gender, days of ventilation and intracranial hemorrhage according to the classification by Papile et al. 9 There was no infant with cystic periventricular leukomalacia.


Follow-up assessments including laboratory investigations of CMV virus culture and CMV DNA PCR in urine, anthropometry, audiology, neurologic examination and neurodevelopmental assessment were conducted at 2 to 4.5 years of age. Age was corrected for gestation up to 2 years. The evaluators were blind to the child’s CMV status.


Body length, body weight and head circumference were measured by a specially trained nurse. We used the standard growth and weight charts by Largo et al., 10 and microcephaly was defined as a head circumference below the second standard deviation. In one CMV-positive boy with cerebral palsy, a correct height measurement was not possible because of contractures; in another CMV-positive boy, body weight could not be obtained because of lack of cooperation.

Laboratory investigations at follow-up.

Urine samples were obtained for investigation of cytomegalovirus culture and CMV DNA PCR analysis. The methods used have been described previously by Hamprecht et al. 3


All 44 infants underwent otoscopic examination and audiometric tests. Hearing sensitivity was assessed by pure tone audiometry with the use of earphones or sound fields. Middle ear function was assessed by impedance audiometry. Transiently evoked otoacoustic emissions were used as an objective method. Sensorineural hearing loss was defined as a pure tone threshold >20 dB without significant air-bone gap.

Neurology and development.

A structured neurologic examination was carried out and the findings were classified as normal, suspect or abnormal. A rating of normal was given for an entirely normal neurologic examination. The following findings were classified as suspect: nonspecific signs like asymmetry of muscular tone or reflexes; muscular hypotonia; or muscular hypertonia (but without definite signs of spastic cerebral palsy as described below). A rating of abnormal was given when there were clear neurologic signs of spastic cerebral palsy (increased muscular tone with pathologic reflexes including pyramidal signs and an abnormal pattern of movement and posture). Subtypes of spastic cerebral palsy were categorized according to the categories used by Kraegeloh-Mann et al. 11 The assessments of motor and speech and language development were based on the concept of essential developmental milestones. According to this concept development is classified as normal when it is within the 90th centile or delayed when it is below the 10th centile for age according to the standards of Largo et al. 12, 13


Statistical analysis was performed with SPSS Version 10.0 for Windows. The χ2 and Fisher exact tests where appropriate were used to analyze categoric data. The paired t test was used for analysis of noncategoric data. p <0.05 was considered significant.


Characteristics of study population and controls.

In the control group 17 mothers were CMV-seronegative and 5 mothers were seropositive, but there was no transmission of CMV to the infant during the first 6 months of life. Median birth weight was 1020 g (range, 600 to 1870 g) in the CMV-positive group and 975 g (range, 490 to 1700 g) in the CMV-negative group. Median of gestational age was 27.6 weeks (range, 23.6 to 32 weeks) in the CMV-positive group and 28.4 weeks (range, 23.6 to 31.7 weeks) in the CMV negative group. Median of duration of mechanical ventilation in both groups was 5 days (range, 0 to 37 days in the CMV-positive group and 0 to 23 days in the CMV-negative group).

Neonatal clinical symptoms, onset and duration of viruria.

Table 1 shows for each patient the time of onset of viruria, neonatal clinical symptoms and CMV DNA viruria (last positive and, where available, first negative day) as well as neurologic and neurodevelopmental status at follow-up. The onset of CMV DNA viruria was between Day 23 and Day 190 postnatally. In five infants viruria was detected at clinical follow-up visits after discharge from the neonatal unit.

Clinical and virologic follow-up including time of onset of CMV DNA viruria, neonatal clinical symptoms: CMV DNA viruria, neurology and developmental status of the CMV-positive infants at follow-up age 2 to 4.5 years

Neonatal clinical symptoms including laboratory findings such as thrombocytopenia and neutropenia associated with CMV infection were present in 12 of the 22 CMV-positive infants (Table 1). Four infants developed sepsis-like symptoms, and 5 infants had signs of hepatopathy. Viruria was detected in 1 infant with hepatitis at a clinical follow-up after discharge 2 days after the onset of clinical symptoms. Petechia were present in 1 infant.

At follow-up at age 2 to 4.5 years, 14 (64%) of the patients continued to excrete CMV DNA in urine. The median of minimum time of viruria was 1090 (range, 95 to 1612) days. In the CMV-negative control group 1 patient, who was negative at the age of 6 months, had CMV viruria at 2.9 years.

Neurology, motor development, speech and language development.

There was no difference between CMV-positive and CMV-negative subjects when comparison was made for neurologic status, motor development and development of speech and language.

Normal neurology was present in 19 infants in the CMV-positive group and in 18 infants in the CMV-negative group. Neurologic status was rated as “suspect” in 2 infants in the CMV-positive and in 3 infants in the CMV-negative group. In each group there was 1 child with bilateral spastic cerebral palsy. The infant in the CMV-positive group had severe 4-limb-dominated bilateral spastic cerebral palsy after a Grade IV intracranial hemorrhage. The onset of viruria in this child was at a chronologic age of 90 days; he did not have sepsis-like symptoms in the neonatal period. The infant in the CMV-negative group had a mild leg-dominated bilateral cerebral palsy and normal cranial ultrasound. In 2 of the 4 infants who showed sepsis-like symptoms in the neonatal period, neurologic status was classified as suspect. Both patients had mild muscular hypotonia and mild delay of motor development. One of them had a Grade II hemorrhage, and the other had normal cranial ultrasound evaluation in the neonatal period. Age-appropriate motor development was present in 19 infants in each group. Motor development was delayed in three children in each group. With regard to speech and language development, a normal status was seen in 14 infants in each group. Speech and language development was delayed in 8 children in each group, among them 1 CMV-positive infant who had SLS in the neonatal period. Table 1 shows the individual neurodevelopmental and neurologic status for the CMV-positive subjects.

Audiometric evaluation.

None of the 44 children had sensorineural hearing loss. Twenty-six children had normal pure tone threshold <20 dB hearing and normal transiently evoked otoacoustic emissions. In 5 children we found a reduced compliance of the tympanic membrane without impairment of hearing. Thirteen children had conductive hearing loss caused by middle ear effusion but normal sensorineural hearing.


At follow-up the anthropometric measurements of the study group did not differ significantly (P > 0.1 for body weight, body length and head circumference) from the control group.

Mean body length in the study group was −0.9 ± 1.3 SD scores (SDS), and that in the control group was −1.2 ± 1.3 SDS below the standard growth charts by Largo et al. for term infants. Compared with the standards body weight was −0.9 ± 1.1 SDS in the study group and −1.4 ± 1.0 SDS in the control group. Mean head circumference in the study group was −1.6 ± 1.1 SDS and −1.3 ± 1.0 SDS in the control group. Microcephaly was present in eight infants of the study group and in six infants of the control group.


In contrast to most of the other studies that investigated the effect of CMV infection in early life, our study consists of a homogeneous group of subjects in terms of gestational age; in addition it was possible to determine timing of infection accurately. No subjects with prenatal CMV infection were included. We have demonstrated that in this group of premature infants with postnatally acquired CMV infection transmitted via breast milk at the age of 2 to 4.5 years, sensorineural hearing, neurologic and neurodevelopmental status as well as growth were not different from those of individually matched CMV-negative infants.

Sepsis-like symptoms in the neonatal period occurred in four infants with CMV infection. In two of those four infants neurologic status was suspect, and a third infant had delay of speech and language development. Because the numbers are small, we were not able to assess the significance of SLS in the neonatal period for long term outcome of these infants.

Body measurements and head circumference were not significantly different between the CMV-positive and CMV-negative infants. Although the sample size was small, our findings suggest that there is no negative effect of early postnatally acquired CMV infection on development of height and weight. Mean body length, body weight and head circumference in both groups were below the standards for term born infants.

Sensorineural hearing loss is the most common consequence of congenital CMV infection. 14 Onset of hearing loss is often delayed (up to the age of 6 years), and the course is frequently progressive and fluctuating. 15 With regard to peri- and postnatally acquired CMV infections, the results of previous studies are less clear. Paryani et al., 8 who examined a group of term born and premature infants at the age of 3 years, and Johnson et al. 16 in a similar study did not find a significant increase of sensorineural hearing loss in infants with postnatally acquired CMV infection. These data are supported by our study in which no infant had sensorineural hearing loss. It must be kept in mind that we examined the infants between 2.5 and 4 years of age, and it will be necessary to perform another follow-up assessment at school age to exclude delayed onset of hearing loss.

We could not detect a difference in neurodevelopmental outcome between the CMV-positive and CMV-negative group. In both groups there was a relatively high frequency of speech and language delay, but there was no difference between the groups. With regard to neuromotor function, in each group there was one child with clearly abnormal neurologic status. Results of earlier studies, for example the study conducted by Paryani et al. in 1985, 8 suggest that, in contrast to term born infants, the risk for neurologic sequelae in very preterm infants with postnatally acquired CMV infection may be increased, especially when the onset of CMV excretion was in the first 2 months of life. In our study, within the CMV-positive group those with neurodevelopmental problems were infants with early excretion of CMV in urine. This is in keeping with the results of the above mentioned study. 8

We used the concept of milestones as a tool for assessment of neurodevelopmental status at the age of 2 to 4.5 years. This method allows reliable categorization of the neurodevelopmental status as either normal or delayed. However, to detect possible subtle neuropsychologic impairments in preterm children with early postnatally acquired CMV infection, it is necessary to undertake a second follow-up assessment at school age including detailed psychometric testing. We are currently planning to recall the subjects of this study for a detailed second follow-up assessment at school age.


This study was partly supported by the Beitlich-Stiftung, Tubingen, Germany.


1. Lawrence RM, Lawrence RA. The evidence for breastfeeding. Pediatr Clin North Am 2001;48:235–51.
2. Schanler RJ. Suitability of human milk for the low birth-weight infant. Clin Perinatol 1995;22:207–22.
3. Hamprecht K, Maschmann J, Vochem M, Dietz K, Speer CP, Jahn G. Epidemiology of transmission of cytomegalovirus from mother to preterm infant by breastfeeding. Lancet 2001;17:513–8.
4. Maschmann J, Hamprecht K, Dietz K, Jahn G, Speer CP. Cytomegalovirus infection of extremely low-birth weight infants via breast milk. Clin Infect Dis 2001;33:1998–2003.
5. Vochem M, Hamprecht K, Jahn G, Speer CP. Transmission of cytomegalovirus to preterm infants through breast milk. Pediatr Infect Dis 1998;17:53–8.
6. Ballard RA, Drew WL, Hufnagle KG, Riedel PA. Acquired cytomegalovirus infection in preterm infants. Am J Dis Child 1979;133:482–5.
7. Dworsky M, Yow M, Stagno S, Pass RF, Alford C. Cytomegalovirus infection of breast milk and transmission in infancy. Pediatrics 1983;72:295–9.
8. Paryani SG, Yeager AS, Hosford-Dunn H, et al. Sequelae of acquired cytomegalovirus infection in premature and sick term infants. J Pediatr 1985;107:451–6.
9. Papile LA, Burstein J, Koffler H. Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1500 gm. J Pediatr 1978;92:529–34.
10. Largo RH, Walli R, Duc G, Fanconi A, Prader A. Evaluation of perinatal growth: presentation of combined intra- and extrauterine growth standards for weight, length and had circumference. Helv Paediatr Acta 1980;35:419–36.
11. Kraegeloh-Mann I, Hagberg G, Meisner C, et al. Bilateral spastic cerebral palsy: a comparative study between southwest Germany and western Sweden. I. Clinical patterns and disabilities. Dev Med Child Neurol 1993;35:1037–47.
12. Largo RH, Molinari L, Comenale Pinto L, Weber M, Duc G. Language development of term and preterm children during the first five years of life. Dev Med Child Neurol. 1986;28:333–50.
13. Largo RH, Molinari L, Weber M, Comenale Pinto L, Duc G. Early development of locomotion; significance of prematurity, cerebral palsy and sex. Dev Med Child Neurol 1985;27:183–91.
14. Hicks T, Fowler KB, Richardson M, Dahle AJ, Adams L, Pass RF. Congenital cytomegalovirus infection and neonatal auditory screening. J Pediatr 1993;123:779–82.
15. Fowler KB, McCollister FP, Dahle AJ, Boppana S, Britt WJ, Pass RF. Progressive and fluctuating sensorineural hearing loss in children with asymptomatic congenital cytomegalovirus infection. J Pediatr 1997;130:624–30.
16. Johnson SJ, Hosford-Dunn H, Paryani SG, Yeager AS, Malachowski N. Prevalence of sensorineural hearing loss in premature and sick term infants with perinatally acquired cytomegalovirus infection. Ear Hear 1986;7:325–7.

Cytomegalovirus infection; preterm; breast milk; postnatal infection; hearing; development

© 2004 Lippincott Williams & Wilkins, Inc.