Skip to main content

Key Research 2012-2013

Key Research 2012-2013


Baddock, S.A., Galland, B.C., Bolton, D.P.G., Williams, S.M. and Taylor, B.J. 2012. “Hypoxic and Hypercapnic Events in Young Infants During Bed-Sharing.” Pediatrics 130(2): 237–44.

Observation of forty healthy fullterm regularly bed-sharing infants from 0 to 6 months old and 40 matched control co-sleeping infants. Infant and parent behaviour, arterial oxygen saturation and carbon dioxide inspiration were recorded.

Objective – in many cultures bed-sharing is the norm and holds particular importance and many mothers end up doing so despite being informed about the potential risks. This study aimed to understand the usual practices of bed-sharing infants in their normal environment to understand the impact of the environment on their physiology as compared to co-sleeping infants.


  • The bed-sharing environment is very different to that of other forms of co-sleeping, and this can be both beneficial and potentially compromising. This low-risk population exercised sufficient homeostatic responses and maternal responses in order to keep infants safe.
  • If an adult in the bed is unresponsive, a smoker, or having consumed alcohol or drugs, it has the potential to be highly dangerous for the infant. Equally, overbedding is very dangerous for bed0sharing infants.
  • Oxygen desaturation events <90% were more common in bed-sharing infants, likely due in part to the warmer microenvironment this creates. However events that cause oxygen desaturation, such as episodic apnoea and periodic breathing is not observed more frequently in future SIDS victims.
  • There was also significantly more rebreathing events in bed-sharing infants. There were 80 episodes of rebreathing between 22 bed-sharing infants but only 1 co-sleeping infant. Almost all these episodes were preceded by head-covering.
  • In response to rebreathing events, infants often increase ventilation effectively (mainly in tidal volume)or remove themselves from the situation or prompt action by the mother. Mothers sometimes spontaneously and reactively remove offending bedding during dangerous and potentially asphyxial head covering
  • Over 70% of desaturation events were preceded by central apnoea, between 5 and 10 seconds in duration without bradycardia. This normally occurred in active sleep.
  • Severe apnoea >15 seconds was rare, but occurred in both groups, with 3 events in bed-sharing infants, 6 events in co-sleeping infants.
  • Severe oxygen desaturation <80% was also rare overall (2%), with 3 events in bed-sharing infants and 4 events in co-sleeping infants.


This study has a large sample size, of 40 bed-sharing infants and 40 more co-sleeping variable matched control infants, with roughly 10 hours of overnight sleep observed per infant. The study was also conducted in the normal home environment that recreated the natural home practices of bed-sharers and co-sleepers. 


Sample population has a low-risk of SIDS, as is common in willing study populations. There were few maternal smokers and a high level of maternal education and breastfeeding rate. The sample is unlikely to fully represent the broad spectrum of bed-sharing mothers, including social deprivation.

Galland et al (2012) Normal sleep patterns in infants and children: A systematic review of observational studies Sleep Medicine Reviews 16(3) 213-222

A systematic review of the scientific literature with the purpose to provide an international average reference point for normal sleep patterns in infants and children (0–12 years). Database searches produced 2114 articles with 34 articles qualifying for review. Mean and variability data for sleep duration, number of night wakings, sleep latency (time it takes to fall asleep), longest sleep period overnight, and number of daytime naps were extracted from questionnaire or diary data.


Analysis provided the following reference values and ranges for sleep duration (hours) were: infant (0–23 months), 12.8 (9.715.9); toddler/preschool (2–5 years), 11.9 (9.913.8); and child (6–12 years), 9.2 (7.610.8).

Sleep duration is the most commonly reported sleep variable. It has a wide range in infancy with the greatest rate of change occurring within the first 6 months of life.

Meta-regression showed predominantly Asian countries had significantly shorter sleep (1 h less over the 0-12 year range) compared to studies from Caucasian/non-Asian countries.

Night waking data provided 4 age-bands up to 2 years ranging from 0 to 3.4 wakes per night for infants (0-2 months), to 02.5 per night (12 year-olds).

Sleep latency data were sparse but estimated to be stable across 0-6 years.

In summary, sleep patterns show developmental trends for sleep duration (decrease 0-12 years), number of night wakings (decrease from 0 to 2 years), longest sleep period (increase from 0 to 2 years), and number of daytime naps (decreasing up to age 2).


As the main data analysis combined data from different countries and cultures, the reference values could be considered as global norms.

The findings show a wide variation in what is considered normal sleep pattern, particularly in the very young. This is an important consideration when dealing with infant/child sleep problems.


The data were collated from parental reports about their child, therefore do not represent the more accurate physiological measures of sleep obtained by polysomnography, actigraphy or videosomnography.

Nearly all the studies included in the review failed to provide a breakdown to breast or bottle-feeding, well known to influence sleep patterns. Other aspects of sleep that had limited documentation; normal sleep patterns in the toddler age group, gender and ethnic differences, and weekday versus weekend differences across all age groups.


Longitudinal studies, rather than cross-sectional studies, are required to provide richer sources of data to document developmental patterns of sleep. The gap in the literature around the toddler age group, suggests this age needs to be targeted to better document normal sleep patterns before children’s daytime routine is changed to fit school schedules. Infant sleep location also requires closer attention.

Oden, R.P. et al., 2012. Swaddling: will it get babies onto their backs for sleep? Clinical pediatrics, 51(3), pp.254–9.

Cross-sectional descriptive survey involving 103 parents or guardians, recruited from a predominantly urban African American community.

Objective – to assess whether swaddling promotes sleeping in the safer supine position in a population with high rates of nonsupine positioning.


  • Most participants found swaddling was effective in comforting the infant (45.0%) and promoting sleep
  • Swaddling may encourage parents who do not normally put their infant to sleep in the supine position to do so. 52.6% of the usually nonsupine sleeping infants slept supine when swaddled.
  • Parents tend to put their infant prone or on their side to sleep because they think the infant sleeps better. On the other hand, unsurprisingly, the parents who routinely swaddle their child find that they sleep better when swaddled.
  • There is a perception amongst parents that swaddling may be dangerous for their infant, particularly regarding breathing difficulties.
  • There are dangers associated with swaddling of which parents were not aware, such as hip dysplasia, risk of overheating or head covering (in some circumstances). When practiced correctly swaddling is very safe.


There is a very detailed breakdown of responses that is very illuminating to a wide range of possible factors involved with routine-swaddlers and non-routine swaddlers.


Survey does not allow for probing and further investigation into respondents’ reasoning. No information about future plans to swaddle, whether they plan to stop swaddling and why. This would be useful as swaddling is dangerous once the baby can roll on its own accord. The study recognises that they are limited by their sample size and geographic distribution. However this is because the study is particularly interested in the attitudes towards swaddling amongst a community with a high frequency of infants sleeping in a nonsupine position. This is in order to understand this perspective so that any future efforts to modify this potentially dangerous cultural practice are more effectively directed.

Randall, B.B. et al., 2013. Potential Asphyxia and Brainstem Abnormalities in Sudden and Unexpected Death in Infants. Pediatrics, 132(6), pp.1616–1625.

Re-examination of sudden infant death in relation to “potential asphyxia”.

Some brainstem abnormalities are associated with deficiencies of medullary neurochemical transmitters, such as serotonin, GABA or 14-3-3 transduction proteins. This study hypothesises there is a relationship between potential asphyxial risk factors in the environment and severity of the brainstem abnormality. Theoretically infants that died without any obvious asphyxia risk factors will have greater neurochemical deficiencies, and likely would have been classified as SIDS. Infants dying with obvious asphyxia risks will have more normative neurochemical parameters.


  • Sample size of infants dying of undisputable asphyxia was insufficient to come to be conclusive.
  • No significant difference in medullary neurochemical abnormalities in infants from both asphyxia generating and asphyxia safe environments.
  • No relationship between asphyxial conditions in the sleep environment and brainstem abnormalities in infants dying unexpectedly.
  • However those infants dying suddenly and unexpectedly had lower measures of neurochemicals than the control group of infants that died of known causes (P<0.05).
  • There was a positive correlation between number of risk factors and serotonin1A receptor binding. In other words, fewer exogenous risk factors are required to precipitate infant death when the serotonin network is deficient.
  • This alteration in the medullary serotonin network could be responsible for protective respiratory and autonomic responses.


The researchers used an unclear and subjective scale to classify likeliness of cases due to asphyxia. This is a difficult factor to make objectively measureable. They quantified the number of factors in each environment, but there is no way to differentiate between the severity of each factor. Equally there is detail lacking in forensic reports of the death scene in which infants were found. As the study was based on data not collected by the authors there was no way to be objective or consistent in observations in each scene. Those recording elements of the scene in which the infant was found dead may not be aware of salient factors.

The two different data sets (2006 and 2010) analysed in this study examined medullae for different neurochemical features.

The dataset is relatively small, comprising 71 infant deaths, which when divided into different levels of asphyxia risk was difficult to create meaningful sample sizes.

Vennemann et al. 2012. Bed Sharing and the Risk of Sudden Infant Death Syndrome: Can We Resolve the Debate? The Journal of Pediatrics. 160(1):44-48

Meta-analysis of data on the relationship between bed-sharing and SIDS risk, based on studies conducted between 1984 and 2006 in New Zealand, California USA, Scotland UK, Chicago USA, Norway, USA, Ireland, Germany and England UK.


  • Found an increased risk of SIDS for infants overall (OR 2.89, CI 1.99-4.18)
  • Risk associated with bed-sharing was greatest for infants of mothers who smoked (OR 6.27, CI 3.94-9.99), and infants aged under 12 weeks (OR 10.37, CI 4.44-24.21)
  • No increased risk for babies over 12 weeks, regardless of smoking status
  • Increased risk of SIDS with bed-sharing for non-routine bed-sharers (OR 2.18, CI 1.45-3.28) but not for routine bed-sharers (OR 1.42, CI 0.85-2.38)


By combining data from multiple studies, a very large sample size has been obtained – 2464 cases and 6495 controls, of which 28.8% and 13.3% bed-shared respectively.


The definition of bed-sharing used in this analysis included sofa-sharing, a sleep location known to be associated with a greatly increased risk of SIDS. Data on feeding method (breast or formula) and drugs/alcohol consumption were not included in this analysis. The former is known to affect the way in which mothers and infants behave during sleep, and the latter to increase the risk of SIDS associated with bed-sharing.


Despite the large sample size achieved here by combining data from multiple studies, the usefulness of the results obtained is limited by the combining of sofa-sharing and bed-sharing, and the omission of analyses based on feeding method and drugs/alcohol consumption.


Huang, Y., Hauck., F.R., Signore, C., Yu, A., Raju, T.N., Huang, T.T. and Fein, S.B. 2013. ‘Influence of bedsharing activity on breastfeeding duration among US mothers’. JAMA Pediatrics. 167(11): 1038:44.

Longitudinal data gathered from the Infant Feeding Practices Study II. This enrolled mothers whilst pregnant and followed them through the first year of infant life, sending out questionnaires at 1-7, 9, 10 and 12 months postpartum. 1,846 mothers answered questions bed-sharing and were breastfeeding their infants at 2 weeks.

Objective – to quantify the association between bed-sharing behaviour and breastfeeding prolongation, hypothesising that the longer the duration of bed-sharing, the longer the duration of any and exclusive breastfeeding.


  • Longer duration of bed-sharing as measured by the cumulative bed-sharing score was associated with longer duration of any breastfeeding, but not exclusive breastfeeding after adjusting for covariates.
  • Breastfeeding duration was also associated with factors other than bed-sharing, such as a higher level of education, being white, previously breastfeeding and planning to breastfeed and not returning to work within the first year postpartum.
  • Strategies to increase breastfeeding rates should not focus only on bed-sharing, but take a more multifaceted approach.


The measures of bed-sharing behaviour included specific information about how the mother slept while lying down with her infant. Bed-sharing was measured before breastfeeding cessation, meaning that directionality of the associations could be inferred. The distribution of the sample was wide, across the United States. There was also a high response to the questionnaire amongst mothers through the postnatal period (66.7%). The sample of mothers was large enough that mothers that are normally underrepresented had enough subjects to be included in the analyses.


White, educated, higher socioeconomic status women were overrepresented, decreasing the generalisability of the conclusions. Duration of bed-sharing was based on number of times each mother indicated she was bed-sharing, which reflects the frequency with which the mother was asked during the study, rather than as a continuous time duration.

Stremler, R., Hodnett, E., Kenton, L., Lee, K., Macfarlane, J., Weiss, S., Weston, J. and Willan, A. 2013. “Infant Sleep Location: Bed Sharing, Room Sharing and Solitary Sleeping at 6 and 12 Weeks Postpartum.” The Open Sleep Journal 6(1): 77–86.

Randomised control trial of 246 healthy-first time mothers. The sleep intervention provided information regarding planned and actual infant sleep location at 6 and 12 weeks postpartum when maternal and infant sleep was measured using actigraphy and a maternal sleep diary. This measured subjective maternal sleep disturbance, breastfeeding exclusivity, fatigue and depressive symptomatology.

Objective – to understand the effects of infant sleep location on infant or maternal sleep and other health outcomes. Due to lack of information regarding the prevalence of infant sleep locations this study also measured the disparity between planned and actual infant sleep location.

It is possible for this disparity to be understood with the results regarding infant and maternal sleep and other health outcomes associated with infant sleep location.


  • Most women planned room-sharing with their infant (65.5%), next common was planning to use both bed-sharing and room-sharing (22%).
  • The most common infant sleep location was room-sharing (46% at 6 weeks and 39% at 12 weeks).
  • Usual bed-sharing was common (17% at 6 weeks and 12% at 12 weeks). Bed-sharing to any extent was quite common at 6 and 12 weeks postpartum, at 50.8% and 41.1% respectively.
  • Whilst rates of room-sharing and bed-sharing declined between 6 and 12 weeks postpartum, generally moving to solitary sleeping arrangements, room-sharing or solitary sleeping infants rarely converted to bed-sharing at 12 weeks.
  • Infant sleep location had effects on maternal sleep continuity but was not related to other outcomes, such as fatigue, maternal depressive symptoms or breastfeeding exclusivity, in contrast to Huang et al.’s (2013) study.
  • Bed-sharing mothers had shorter stretches of sleep than solitary sleeping mothers at 6 weeks (130 mins vs 156 mins; p=0.03).
  • Bed-sharing mothers also had more awakenings than those usually room-sharing and solitary sleeping (11.2 vs 8.9 vs 8.3; p=0.001)
  • At 12 weeks mothers who were usually room-sharing had shorter stretches of sleep than those usually solitary sleeping (164 vs 192 mins; p=0.04). There were no significant differences between infant sleep location groups on any other outcomes.


Intention for solitary sleeping was not asked for, instead those who were not bed-sharing or room-sharing were assumed to intend to sleep solitarily. Classifying families as either room-sharing or bed-sharing is also problematic, as families tend to lie somewhere on a spectrum between the categories. The sample was also predominantly highly educated, partnered and of a higher socioeconomic status overall. The findings would be more generalisable if they were more reflective of a greater range of people as bed-sharing tends to be more common in women without partners and with a lower socioeconomic status.