Review Article

Is “Bed Sharing” Beneficial and Safe during Infancy? A Systematic Review

Table 2

Observational studies on bed share and SIDS.

S. no.Study ID/countryDesign, settingStudy populationNumber of subjectsIntervention/exposureOutcomeComments

1Klonoff-Cohen and Edelstein 1995/USA [19]Case-control study, population basedCases: all infants from birth to 1 year of age died of SIDS; controls: matched for birth hospital, sex, race, date of birth, and the same surveyCases, 200
Controls, 200
BedshareSIDS rate = 60/200 (case), 52/200 (control)
No separate data for neonates. The study reported the odds for bed sharing during the daytime also (we did not use that data). OR adjusted for passive smoking. Loss to followup was 25%

2Brooke et al. 1997/Scotland [20]Case-control study, population based Cases: all infant deaths occurring from 7th day of life to 1 year; controls: births immediately before and after the index case in the same maternity unit and matched for the same surveyCases, 146
Controls, 275
Bed-share
SIDS rate = 11/146 (case), 6/275 (control)
No separate data for neonates available. Other factors studied, GA ≤ 36 wks, BW < 2500 g. Loss to followup was 25%

3L’Hoir et al. 1998/The Netherlands [21]Case-control study, population basedCases: all sudden deaths from 7 days to 2 years of age; controls: matched for date of birth and the same survey Cases, 73
Controls, 146
Bed-share
SIDS rate = 6/73 (case), 7/146 (control)
Out of 73 SIDS, only 10 happened during the neonatal period. Other factors studied: smoking, alcohol. Data provided for infants who were not exposed to passive smoking

4Blair et al. 1999/United Kingdom [22]Case-control study, population basedCases: all unexpected deaths up to 2 years of age; controls: infants born immediately before and after the index case and matched for the same surveyCases, 321
Controls, 1299
Bed-share
SIDS rate = 82/321 (case), 189/1299 (control)Separate data for neonates not available. Other factors studied: smoking, alcohol. Only 23 infants died between 7 and 60 days of life

5Arnestad et al. 2001/Norway [23]Case-control study, population basedCases: all sudden deaths among children between the 2nd week and 3 yrs of age; controls: infants matched for sex and date of birth, randomly picked from the national register and matched for the same surveyCases, 174
Controls, 375
Bed-share
SIDS rate = 15/174 (case), 24/375 (control) No separate data for neonates available (out of 174 cases, only 13 died before 2 months of age). Other factors studied: smoking, breast feeding, birth order and weight, mode of sleeping, dummy use, and socioeconomic factors. Adjusted for passive smoking. Loss to followup was 31% (case) and 25% (control)

6Williams et al. 2002/New Zealand [24]Case-control study, population basedCases: all unexpected infant deaths from 29 days to 1 year of age; controls: randomly selected from all births, except home births, and matched for the same surveyCases, 369
Controls, 1558
Bed-share
SIDS rate = 86/369 (case), 162/1558 (control)No separate data for neonates available. Other factors studied: smoking, breast feeding. Bed sharing refers to “usual” pattern or last night’s sleep is not known. Loss to followup was 10–19%

7Carpenter et al. 2004/Europe [25]Case-control study, population basedCases: all unexplained deaths in the first year of life; controls: randomly selected from the birth records, matched for age and the same surveyCases, 281
Controls, 1760
Bed-share SIDS rate = 32/281 (case), 139/1760 (control)
Of the total 700 odd cases, only 57 SIDS occurred in the first month of life. Other factors studied: smoking, alcohol. Information depicted here is for infants whose mothers did not smoke

8Bubnaitiene et al. 2005/Lithuania [26]Case-control study, population basedCases: included <1 year of age group died of SIDS; controls: matched for date of birth, region, and the same surveyCases, 35
Controls, 145
Bed-share
SIDS rate = 0/35 (case), 20/145 (control)
No separate data for neonates available (only 1 SIDS during the neonatal period). Studied subgroups, GA ≤ 36 wks, BW <2 500 g. Loss to followup was 22.2% in cases

9McGarvey et al. 2006/Ireland [27]Case-control study, population basedCases: all infants from birth to 1 year of age died of SIDS; controls: matched for date of birth, community area, and the same surveyCases, 259
Controls, 829
Bed-share
SIDS rate = 128/259 (case), 101/829 (control)
Data provided is for infants aged <10 weeks of age. Other factors studied: smoking, alcohol. Loss to followup was 14%

10Ruys et al. 2007/The Netherlands [28]Case-control study, population basedCases: all infants <6 months of age died of cot deaths; controls: infants of the same age groups who participated in a countrywide surveyCases, 138
Controls, 1628
Bed-share
SIDS rate = 36/138 (case), 151/1628 (control)
No separate data for neonates available. Other factors studied: smoking, breast feeding. Adjusted for breast feeding, age, and passive smoking

11Blair et al. 2009/England [9]Case-control study, population basedCases: all unexpected deaths up to 2 years of age; controls: from the maternity database of one hospital and matched for the same surveyCases, 79
Controls, 87
Bed-share
SIDS rate = 30/79 (case), 17/87 (control)
Other factors studied: smoking, narcotics, GA ≤37 wks, BW < 2500 g. Neonates accounted for only 15% of SIDS. Loss to followup was 5–14% in both the groups

12Vennemann et al. 2009/Germany [29]Case-control study, population basedCases: all infants from birth to 1 year of age died of SIDS; controls: matched for date of birth and the same surveyCases, 333
Controls, 998
Bed-share SIDS rate = 27/333 (case), 28/998 (control)
Data provided is for infants aged <13 weeks of age. Other factors studied: smoking, GA ≤ 37 wks, BW < 1500 g. Adjusted for maternal smoking. Loss to followup was 18–42% in both the groups

13Fu et al. 2010/USA [30]Case-control study, population basedCases: all infants from birth to 1 year of age died of SIDS: controls: matched for birth, race, age, birth weight, and the same survey Cases, 195
Controls, 194
Bed-share
SIDS rate = 15/195 (case), 6/194 (control)
Reported data for 3 subgroups: <1 mo, 1–3 mo, and >4 mo; only the 1st month data has been used here. Other factors studied: smoking, alcohol. Bed sharing included sleeping on the mattress as well as sofa. Loss to followup was 25%