Family doctor


Children's Health



Sudden infant death is a tragedy of which the exact cause is unknown. Much can be done to reduce the risk. World expert Associate Professor Ed Mitchell outlines the current understanding and research.

What is SIDS?

Sudden Infant Death Syndrome (SIDS or cot death) is defined as "The sudden death of an infant or young child, which is unexpected by history, and in which a full postmortem examination fails to demonstrate an adequate cause of death."

SIDS can occur anywhere - not just in the cot. Babies can die suddenly and unexpectedly in prams, cars, beds or even someone's arms.

What causes SIDS?

Although we don't know for certain, the most likely explanation is that a combination of factors affects a baby at a critical stage of development. The baby must be predisposed, such as having an arousal defect, which may have been caused by the effects of tobacco smoke when in the uterus. And finally the baby needs to be exposed to some stress, such as prone (front) sleeping position.

How can SIDS be reduced?

The current advice of the New Zealand Cot Death Association to help prevent SIDS is:

  1. Sleep babies on their backs
  2. Have a smoke-free pregnancy and smoke free home
  3. If you smoke or smoked during pregnancy, sleep your baby in its own bed
  4. Breastfeed, if possible

What is the evidence for this advice?

Sleep position

Epidemiological studies in several countries have identified prone sleeping position as an important risk factor for SIDS. The most compelling evidence that prone sleeping position (front sleeping) is causally related to SIDS comes from the major reduction in SIDS rates in the last decade in many countries following prevention programmes changing infants from the prone sleeping position onto their sides or backs.

More recent studies have shown that the risk of SIDS with the side sleeping position is about twice that of infants sleeping on their back, probably because the side sleeping position is relatively unstable, and some babies turn to the prone position.

The major concern about the back sleeping position was the perceived danger of aspiration of vomit. Deaths from such causes are rare in countries that traditionally use the back sleeping position. There has been no increase in deaths from aspiration since the change in sleep position.

The only babies we do not recommend back sleeping position are those with congenital abnormalities of the jaw, such as the Pierre Robin anomaly.


There is substantial evidence to conclude that maternal smoking caused a marked increase in SIDS. There have been almost 50 studies that have examined this relationship and all indicate an increased risk. Since the reduction in the prevalence of prone sleeping position there have been eight studies examining maternal smoking and SIDS. The pooled unadjusted relative risk (RR) from these studies is almost five, which suggests that infants of mothers that smoke have almost a five times risk of SIDS compared with infants of mothers who do not smoke. Adjustment for potential confounders lowers the risk estimate, however, many studies over adjust, such as controlling for birthweight, resulting in an inappropriate low estimate of the risk.

Epidemiologically it is difficult to distinguish the effect of active maternal smoking during pregnancy from involuntary postnatal tobacco smoking of the infant to smoking by the mother. The mechanism for SIDS is unknown, however, it is generally believed that the predominant effect from maternal smoking is from in-utero exposure of the fetus.

Clear evidence for environmental tobacco smoke exposure can be obtained by examining the risk of SIDS from paternal smoking where the mother is non-smoker. There have been six such studies. The pooled unadjusted RR was 1.4, which is much smaller than the effect seen for maternal smoking which was 4.7.

Bed sharing

Since biblical times maternal overlaying of the infant while bed sharing has been thought to be a cause of sudden infant death. Several earlier studies reported an increased risk of SIDS with bed sharing, which was confirmed by preliminary findings in the New Zealand Cot Death Study. Further analysis showed that the increased SIDS risk from bed sharing was predominantly among infants of mothers who smoked and this was consistent in all three main ethnic groups.

A recent meta-analysis found that the pooled relative risk for bed sharing was 2.1 among infants of mothers who smoked, but only 1.4 for infants of mothers who did not smoke. We have recommended that public health policy should be directed against bed sharing by infants whose mothers smoke, as they carry an increased risk of SIDS from bed sharing in addition to their already increased risk from maternal smoking. For infants of nonsmoking mothers, who have a low absolute risk of SIDS, the 40-50% increase in risk needs to be balanced against other perceived benefits from bed sharing, such as increased breastfeeding.

There is no evidence of any group that is at lower risk of SIDS when bed sharing.


There is not uniform agreement that breastfeeding reduces the risk of SIDS. The reduction in SIDS risk is relatively modest, so that a clear beneficial effect is only seen with large studies, such as the NICHD Cooperative Epidemiological Study of SIDS Risk Factors and the New Zealand Cot Death Study. Smaller studies such as that from the United Kingdom found a protective effect with breastfeeding, but after socioeconomic adjustment the effect was no longer significant.

Because there was a clear beneficial effect in the New Zealand Cot Death Study, breastfeeding continues to be part of the New Zealand SIDS prevention programme.

Other factors

Head covering

Approximately 15% of SIDS deaths are found with their heads covered by bedding. It is not certain whether this occurs as a terminal event or whether this causes the death by thermal stress or rebreathing of expired gases. There is some evidence that tucking the baby in securely is associated with a reduced risk of SIDS, probably by preventing the bedding riding up over baby's face. Although there is no scientific evidence supporting the UK recommendation that infants should be placed at the foot of the cot to reduce the risk of head covering, it is unlikely to be harmful.

Thermal factors

Thermal factors, such as excess bedding, illness, swaddling, season and latitude, were associated with an increased risk of SIDS. These factors appeared to be only important for infants sleeping prone. Few infants now sleep prone. The decline in SIDS has produced two unexpected changes in the epidemiology of SIDS. Firstly, the winter peak has largely disappeared and secondly, the increased SIDS incidence in the colder southern New Zealand, compared with the warmer north, has disappeared.

Thermal factors no longer appear important as has been confirmed in a recent published study from the UK where few babies now sleep prone.

Pacifiers (dummies)

An intriguing observation has been the protective effect of pacifiers. This finding has been confirmed in several studies from different countries. Pacifiers seem to reduce the risk of SIDS by about a half. As pacifiers stay in an infant's mouth for a very short period of time after the infant falls asleep, it is unlikely that they have a direct mechanical effect, but may affect the muscles and tone of the airway.

However, there are detrimental affects of pacifiers. They have been shown to decrease breastfeeding and increase the incidence of otitis media.

At the moment pacifiers are not recommended for SIDS prevention, but conversely, they are not being discouraged.

What is plagiocephaly and how might it be prevented?

There has been an increase in plagiocephaly or flattening of the skull in the last few years. In part this is explained by better recognition, but in part is due to babies lying on their backs. In most instances the flattening is minor and corrects with age. However, some require treatment with helmets to push the skull back into shape for cosmetic reasons. There is no damage to the underlying brain and it does not affect the child's development.

There are some possible ways that plagiocephaly might be avoided. First, when awake the baby can be on his tummy (tummy time). This takes the pressure off the back of the skull. As much of the time the baby sleeps on his back, turning the head to alternate sides may reduce the pressure. Turning the cot round and/or placing favourite toys on alternative sides of the cot might encourage the baby to look the other way and fall asleep in this position.

What babies are at increased risk of SIDS?

There are several groups of infants at higher risk. These include:

  • Maori babies. Much of the increase risk in Maori babies can be explained by the higher rates of maternal smoking and bed sharing.
  • Boys are at higher risk than girls (60:40). The reason is not known.
  • Infants who are born pre-term or small are also at increased risk.
  • Twins
  • Low socioeconomic groups, in part explained by higher smoking rates.

Infants aged one to three months are at highest risk of SIDS. More than 80% of all SIDS cases occur before 6 months of age. Each year in New Zealand there is on average just 2 SIDS deaths after the first birthday.

Is SIDS on the decline?

The table gives the number (and rate per 1000 live births) of all infant deaths (from birth through to the first birthday) and all deaths from SIDS.

Total infant deaths from all causes are shown first followed by deaths from SIDS, including the number and rates.
















































































1997 (provisional)





1998 (provisional)





There was a big fall in mortality between 1989 and 1990, and then a more gradual decline.

What caused the decline?

The fall in mortality between 1989 and 1990 was due to the change in infant sleeping position. In early 1990 it was recommended to place babies to sleep on their side or back, although sleep position began to change in 1988. The more recent slow decline is probably due to more infants sleeping on their back rather than the side.

The provisional figures for 1998 need to be interpreted with caution, as there is evidence of diagnostic shift (e.g. SIDS being classified as mechanical asphyxia or suffocation).

Does toxic gas cause SIDS?

In 1989 Richardson postulated that the cause of sudden infant death syndrome (SIDS) was poisoning by arsines, stibines or phosphine. According to the hypothesis these gases are produced by the fungus Scopulariopsis brevicaulis metabolising chemicals containing arsenic, antimony and phosphorus, which were present as plasticisers or fire retardants in polyvinyl chloride (PVC) covered cot mattresses.

In New Zealand most cot mattresses are made from a single block of foam covered with fabric. The use of PVC covered mattresses is very uncommon, and thus the toxic gas theory did not appear to be relevant. However, it has been suggested that in New Zealand the main cause of SIDS is the biogeneration of phosphine from phosphorus in foam cot mattresses and woven fabrics mattress covers and arsines from arsenic in sheepskins. To prevent these gases reaching the infant, polythene wrapping of cot mattresses, and sheepskins and also adult mattresses, if the infant slept in the same bed as the parent, was recommended.

In May 1998 the Independent Expert Group to investigate cot death theories concluded that the toxic gas theory is unsubstantiated. Their conclusions were based on:

  1. Cot mattress contamination with the fungus S. brevicaulis was rare, and no more common in SIDS mattresses than in other used mattresses.
  2. There was no evidence for biovolatisation of phosphorus, arsenic and antimony from PVC cot mattress samples by S. brevicaulis, under conditions relevant to an infant's cot.
  3. There was no evidence of poisoning by phosphine, arsine and stibine or their methylated derivatives. In particular they noted the absence of features normally associated with arsine/stibine or phosphine poisoning.
  4. Low amounts of antimony could be detected in samples from the majority of live infants, and the concentrations in SIDS infants are not exceptional. The presence of antimony in fetal tissue suggested maternal transfer during pregnancy.
  5. There was no evidence that the changing SIDS rates corresponded to the introduction and removal of antimony and phosphorus containing fire retardants in cot mattresses. Although some features of the epidemiology of SIDS is compatible with the hypothesis, a detailed review of the epidemiological evidence does not support it.

The recommendation to wrap cot mattresses with polythene does not appear to be protective as three SIDS deaths have occurred on polythene wrapped cot mattresses in the United Kingdom. Furthermore, the use of thin polythene increases the risk of suffocation.

The Ministry of Health states: "The weight of currently available evidence is that the use of impermeable mattress covers, such as ones made of PVC, polythene and rubber, neither increases nor decreases the risk of SIDS. There is, however, definite evidence that plastic sheeting, including plastic bags, in a baby's sleeping environment has caused the death of babies through suffocation, although the number of reported deaths from this cause is currently very low."

Does immunisations cause SIDS?

The age distribution of SIDS coincides with the timing of immunisations of infants, which has led to the suggestion that immunisations may be the cause of some deaths from SIDS. Several studies have examined the relationship between SIDS and immunisation. Some early studies supported this contention, but more recent studies, including one from New Zealand, have shown that immunisation is associated with either no increased risk or even a lower risk of SIDS.

What is the role of apnoea monitors?

Apparent life threatening events (ALTE) is where a baby is found pale, cyanosed or not breathing and is thought to be dying. In New Zealand 0.7% of infants are admitted to hospital for ALTE. Five percent of healthy babies are reported to have stopped breathing for more than 20 seconds.

Reviews of SIDS cases suggest that 10% have had episodes, which could be interpreted as ALTEs. These need to be interpreted with caution as the information is collected retrospectively and is subject to recall bias. Even if there is a causal association between ALTEs and SIDS, ALTE would only account for about 9% of SIDS cases and that over 500 ALTE cases would need to be treated to prevent one SIDS death. Thus one might conclude that the relationship between ALTE and SIDS is tenuous.

Despite that ALTE is a serious paediatric problem, which causes considerable anxiety to families. A recent working party recommended the following infants should be monitored:

Medical indications
  • Infants with chronic lung disease going home on oxygen
  • Infants with apnoea of prematurity persisting past 35 weeks
  • Infants with Pierre Robin syndrome and other similar abnormalities
  • Infants with ALTE resulting in desaturation or obstruction
Non-medical indications
  • Subsequent siblings of infants who have died of SIDS
  • Surviving twin

*** Hot off the press ***

A recent study has found that changing an infant's sleeping position increases the risk of SIDS.

Infants placed supine (back) to sleep were at the lowest risk of SIDS, which supports the recommendation that this is the preferred sleeping position for healthy infants. The risk was particularly high among infants who usually slept non-prone (side or back), but for some reason were placed prone for the last sleep (adjusted overall risk = 19.3). These infants accounted for 8% of all SIDS deaths.

We recommend extreme caution when placing infants prone if they are used to the non-prone sleeping position. This is particularly relevant for infants in New Zealand where parents have been encouraged to expose infants to supervised prone positioning when awake ("tummy time"). Infants should not be left in the prone position if they fall asleep while playing in the prone position.

In addition, 12% of deaths were infants, who were placed non-prone, but were found prone, that is they had turned during sleep for what may have been the first time. We speculate that experience of the prone position when awake might reduce this risk.

Finally, the study suggested that infants who have established a pattern of spontaneously changing sleep position, including those who turn to prone, are not at increased risk of SIDS.

Reference: Mitchell EA, Thach BT, Thompson JMD, Williams S. Changing infants' sleep position increases risk of sudden infant death syndrome. Arch Pediatr Adolesc Med 1999;153:1136-1141

See also:

Did this article meet your requirements/expectations?