Probiotics for Prevention of Necrotizing Enterocolitis

A review of the effect of probiotics on necrotizing enterocolitis in premature babies was conducted by researchers in The Cochrane Collaboration.  After searching for all relevant studies, they found nine studies done by other researchers.  Their findings are summarised below. 

Please note that this Cochrane review has been updated since the summary below. See the abstract of that update here.

What is necrotizing enterocolitis and why probiotics?

Necrotizing enterocolitis (NEC) is a serious bowel disease. NEC happens when the bowel (intestine) becomes infected and parts of the bowel die.

NEC mostly affects premature babies, and particularly those that are of very low weight (less than 1500 g). Up to 10 out of 100 babies of very low birth weight may develop NEC.

It is not entirely clear what causes NEC. In premature babies, the bowels are not fully developed and are more likely to get infections. It is also possible that the use of infant formula may play a part as babies who are breast fed are less likely to suffer from NEC.

Babies with NEC may have the same symptoms as babies with other digestive diseases, such as problems feeding, vomiting, and bloated and tender stomachs. NEC is usually diagnosed with X-rays and blood tests. If a baby is suspected of having NEC, antibiotics are given, feedings are stopped, and the baby is only fed intravenously (through a vein) until the bowel heals.

In about one third of affected babies, the damage to the bowel is so serious that it perforates (tears). If this happens, bacteria from the bowel can leak into the stomach and cause an infection. At worst, the baby could die. If the bowel perforates surgery is carried out and the damaged parts of the bowel are removed.

Probiotics are dietary supplements that contain potentially "good" bacteria. The most common type of probiotic supplements are yoghurt or other dairy products where these bacteria already exist or where they are added.

The most frequently used probiotics are Lactobacillus and Bifidobacterium. These bacteria already live in the stomach and bowel but premature babies may not have enough of these bacteria. The idea behind feeding the babies probiotics before they become ill is that this may help decrease the amount of "harm" bacteria in their bowel.

What does the research say?

Not all research provides the same quality of evidence. The higher the quality, the more certain we are about what the research says about an effect. The words will (high quality evidence), probably (moderate quality evidence) ormay (low quality evidence) describe how certain we are about the effect.

The studies showed that giving premature babies probiotics:

  • probably prevents severe NEC
  • probably makes little or no difference to the number of babies who develop severe blood infections
  • probably leads to fewer babies dying during their first month

In general, side effects are poorly documented and it is difficult to provide precise information. There is a fear that probiotics may cause infections. In these studies, these side effects were not seen.

The babies in these studies all weighed less than 2500 g. Some of the babies weighed less than 1500 g, while some weighed less than 1000 g. The studies do not show if the treatment was equally effective for all of these groups.

Table of results

What was measured  

No probiotics 


Quality of evidence 

Severe NEC

6 per 100  

2 per 100  

(1 to 3 per 100)1  


Severe infections of all types contracted after birth

15 per 100  

14 per 100 

(11 to 18 per 100)1  


Babies dying from any cause in the first month

6 per 100  

3 per 100  

(2 to 5 per 100)1  


Side effects

No side effects seen in these studies


1The numbers in the brackets show the range in which the actual effect could be.

Where does this information come from?

The Cochrane Collaboration is an independent global network of volunteers, dedicated to summarizing research about health care.

This information is taken from this Cochrane Review: Al Faleh K, Bassler D. Probiotics for prevention of necrotizing enterocolitis in preterm infants. Cochrane Database of Systematic Reviews 2008, Issue 1. Art. No.: CD005496. DOI: 10.1002/14651858.CD005496.pub2.

This summary was prepared by:

Claire Glenton and Gunn Vist, the Nordic Cochrane Centre's Norwegian branch, Norwegian Knowledge Centre for Health Services, on behalf of the Cochrane Complementary and Alternative Medicine Field, and with funding from the US National Center for Complementary and Alternative Medicine (NCCAM) of the US National Institutes of Health (grants number R24 AT001293).

Probiotics for prevention of necrotizing enterocolitis in preterm infants

Patient or population: patients with prevention of necrotizing enterocolitis in preterm infants
Settings: Hospitals in Greece, Israel, Italy (2), Japan (2), Taiwan, Uk and US
Intervention: probiotics


Illustrative comparative risks* (95% CI)  

Relative effect
(95% CI)

No of Participants

Quality of the evidence

Assumed risk 

Corresponding risk 

Severe NEC (necrotizing enterocolitis)
stage II or more by Bell's criteria, diagnosed prior to discarge

59 per 1000  

19 per 1000
(10 to 35)

RR 0.32 
(0.17 to 0.60)



Nosocomial sepsis
positive blood or cerebrospinal fluid cultures taken beyond 5 days of age

152 per 1000  

141 per 1000
(111 to 181)

RR 0.93 
(0.73 to 1.19)



All cause neonatal mortality  

64 per 1000  

28 per 1000
(16 to 48)

RR 0.43 
(0.25 to 0.75)



Any NEC  

See comment

See comment

Not estimable


See comment

Systemic infection with the supplemented organism  

See comment

See comment

Not estimable


See comment

Duration of total parenteral nutrition  

The mean duration of total parenteral nutrition ranged across control groups from 
13.9 to 14.7 days  

The mean Duration of total parenteral nutrition in the intervention groups was
(4.6 lower to 1.9 higher)



very low4,5,6  

Time to establish full enteral feeds  

See comment

See comment

Not estimable


See comment

Duration of hospitalization  

The mean duration of hospitalization ranged across control groups from 
39 to 47 days  

The mean Duration of hospitalization in the intervention groups was
(12 lower to 53 higher)



very low5,6,7  

Neurodevelopmental impairement  

See comment

See comment

Not estimable


See comment

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. Thecorresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence

High quality: Further research is very unlikely to change our confidence in the estimate of effect. 

Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.

Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.

Very low quality: We are very uncertain about the estimate.

1 3/5 unclear randomization procedure, 3/5 unclear allocation concealement, 2/5 unclear blinding, 1 study unclear losses to follow up
2 4/6 unclear randomization procedure, 4/6 unclear allocation concealement, 3/6 unclear blinding, 1 study unclear losses to follow up
3 2 unclear and 1 inadequate randomization procedure, 2 unclear and 1 inadequate allocation concealement, 1 unclear blinding and 1 unclear losses to follow up
4 Unclear allocation ceoncealement in one of the two studies
5 The two studies are so heterogeneous in results that review authors did not combine in metaanalysis
6 Wide confidence intervals
7 Inadequate randomization procedure and allocation concealment in one of the two studies