PEDIATRICS Vol. 108 No. 5 November 2001, pp. 1239-1240

Surprised by Publication

To the Editor.

Before the 20th century, most infants slept in a bed with their parents. An infant who was suddenly and unexpectedly found dead in such an environment was presumed to have been overlain. More recently, with lack of evidence after a thorough postmortem investigation (autopsy), these deaths have been diagnosed as unexplained or sudden infant death syndrome (SIDS). According to the conclusions of Carroll-Pankhurst and Mortimer,¹ we have apparently come full circle and should attribute the sudden death of younger infants to the proximity of larger parents.

Their study design has some fundamental flaws, not least of which is that the data on which their hypothesis is being tested is the same data from which their hypothesis was generated. The authors' findings were based on 26 co-sleeping SIDS infants further split into 2 groups by maternal pre-gravid weight (using arbitrarily defined cutoffs rather than standardized body mass index). It is highly dubious as to whether parametric testing was appropriate, given such small numbers, but more importantly the study used no controls. Whether maternal weight is associated with co-sleeping in the rest of the Cleveland population is therefore unknown. Breastfeeding women commonly bedshare,^2,3 but larger women who breastfeed tend to stop earlier,^4-6 so one might expect bed-sharing by larger women to be associated primarily with younger infants. The larger mothers in this study also had larger babies. Given that the proportional difference in weight between the average mother and baby is in excess of 10-fold, would an increase in maternal weight really elevate the risk of overlaying? To properly assess this issue in relation to bed-sharing, it would be necessary to examine known overlaying deaths (not presumed) with population controls utilizing maternal body mass index for this last sleep.

The observation that bed-sharing SIDS infants are younger is not new and confirms findings from our UK study that the characteristic 3-month peak age of SIDS is actually bimodal, at around 8 weeks for bed-sharers and 15 weeks for those who died in a crib.⁷ This is not surprising, given the higher prevalence of bed-sharing among younger infants.^8-9 What is surprising, if the co-sleeping deaths were attributable to overlaying, is that so few occurred in the first month of life, when bed-sharing is most frequent and the smaller infants presumably were more vulnerable. The authors' criticism of our study was also misplaced; if they had read the references cited,¹⁰ our case ascertainment was 98.3% in a population of 470 000 births.

To conclude, in the absence of any information collected by parental interview or death scene assessment (as strongly recommended recently by the American Academy of Pediatrics in this journal¹¹), that these deaths were probably accidental is irresponsible supposition. To further advocate an avoidance of bed-sharing when co-sleeping deaths on a sofa have been included in the analysis and the contribution of known SIDS risk factors (eg, sleeping position, heavy wrapping, etc) have been ignored is tantamount to the same presumptions made a century earlier.

Peter S. Blair, PhD
Peter Fleming, MD, PhD, FRCP, FRCPCH
Royal Hospital for Children
Bristol BS2 8BJ, United Kingdom

Helen L. Ball, PhD
Parent-Infant Sleep Lab, Department of Anthropology
University of Durham
Durham DH1 3HN, United Kingdom

Martin W. Platt, MD, FRCP, FRCPH
Newcastle Neonatal Service, Ward 35
Royal Victoria Infirmary
Newcastle upon Tyne NE1 4LP, United Kingdom

REFERENCES

1. Carroll-Pankhurst C, Mortimer A Sudden infant death syndrome, bed-sharing, parental weight, and age at death. Pediatrics. 2001; 107:530-536 [Abstract/Free Full Text]
2. Ball HL, Hooker E, Kelly PJ Where will the baby sleep? Attitudes and practices of new and experienced parents regarding cosleeping with their newborn infants. Am Anthropol 1999; 101:143-151 [CrossRef]
3. Clements MS, Mitchell EA, Wright SP, Influences on breastfeeding in southeast England. Acta Pediatr. 1997; 86:51-56 [Medline]
4. Hilson JA, Rasmussen KM, Kjolhede CL Maternal obesity and breastfeeding success in a rural population of white women. Am J Clin Nutr. 1997; 66:1371-1378 [Abstract/Free Full Text]
5. Rutishauser IHE, Carlin JB Body mass index and duration of breastfeeding: a survival analysis during the first six months of life. Breastfeeding Rev. 1993; 2:326-333
6. Donath SM, Amir LH Does maternal obesity adversely affect breastfeeding initiation and duration? J Paediatr Child Health 2000; 36:482-486 [CrossRef][Medline]
7. Blair PS, Fleming PJ, Smith IJ, Ward Platt M, Young J, Nadin P, Berry PJ, Golding J, and the CESDI SUDI Research Group Babies sleeping with parents: case-control study of factors influencing the risk of sudden infant death syndrome. BMJ. 1999; 319:1457-1462 [Abstract/Free Full Text]
8. Tuony PG, Smale P, Clements M Ethnic differences in parent/infant co-sleeping practices in New Zealand. N Z Med J. 1998; 111:364-366 [Medline]
9. Rigda RS, McMillen IC, Buckley P Bed sharing patterns in a cohort of Australian infants during the first six months after birth. J Paediatr Child Health. 2000; 36:117-121 [CrossRef][Medline]
10. Leach CEA, Blair PS, Fleming PJ, Smith IJ, Ward Platt M, Berry PJ, Golding J. Sudden unexpected deaths in infancy: similarities and differences in the epidemiology of SIDS and explained deaths. Pediatrics. 1999;104(4). Available at: http://www.pediatrics.org/cgi/content/full/104/4/e43
11. American Academy of Pediatrics, Committee on Child Abuse and Neglect Distinguishing sudden infant death syndrome from child abuse fatalities. Pediatrics. 2001; 107:437-441 [Abstract/Free Full Text]

To the Editor.

I read with great interest the article by Carroll-Pankhurst and Mortimer,¹ and applaud the authors for an outstanding study. I was concerned, however, of a very important issue in the incidence of SIDS as well as the incidence of bedsharing that is not mentioned: the mode of feeding. Mothers who breastfeed tend to bring their babies to bed, and therefore the incidence of bedsharing is higher among breastfeeding mothers. The data on the incidence of SIDS in breastfeeding versus bottle-feeding infants provide the notion that the incidence is less among breastfeeding children. Do the authors have access to this information?

Ruth A. Lawrence, MD
University of Rochester Medical Center
Department of Pediatrics
Rochester, NY 14642

REFERENCE

1. Carroll-Pankhurst C, Mortimer A Sudden infant death syndrome, bed-sharing, parental weight, and age at death. Pediatrics. 2001; 107:530-536

In Reply.

We appreciate the comments of Blair et al but must disagree with some of them. Our report repeatedly states that we believe bed-sharing to be a cause of only a portion of SIDS-like deaths, and therefore we have not gone full circle as they contend. Rather, our position calls for a compromise between the all-or-none positions that have characterized this debate in the past. If it is correct that our observation explains some undetermined number of such deaths, it may help investigators separate explainable causes from true idiopathic SIDS.

As to the absence of true controls, there have been many classic descriptive epidemiologic studies that demonstrate association and/or causation that did not include controls in the usual sense.¹ Among these are the Framingham studies of precursors of adult cardiovascular disease,² the longitudinal British studies of birth cohorts,³ and the well-known Cleveland family studies.⁴ Moreover, each of these developed hypotheses and tested them from the same data.

In relation to the comment about the paucity of SIDS in the first month of life when bed-sharing is most frequent, we do not know the prevalence of that sleeping practice with young infants in Cleveland. Further, if this concern is valid, should it not apply to the prone versus supine problem as well?

We did not suggest that the report of Blair et al⁵ suffered from underascertainment; instead, we were concerned about the criteria for explained deaths. Their subsequent report⁶ indicates that our concern was unfounded.

Finally, we believe that the bimodal peak age distribution of SIDS (8 weeks for bed-sharers and 15 weeks for crib sleepers) demonstrated in the British studies⁵ supports our concern about co-sleeping. Their implication that this distribution might well explain our results ignores the maternal weight finding. As we stated in our report, each of the factors (bed-sharing, age at death, and maternal weight) is meaningless when considered individually. What is important is the combination of the 3 factors, which led to our conclusion and from which we are not dissuaded. The fact that the significance of this difference disappears when the British data are adjusted for fatigue and alcohol use by the bed-sharing adult, overcrowding in the household, and duvet use does not invalidate this finding. The first 3 characteristics are not risk factors for SIDS, but instead may increase the likelihood of bed-sharing. A basic principle of multivariate analysis in epidemiology is that adjustments for potential confounders are made only for other known risk factors.

In short, we believe that the major criticisms of our data and our conclusions are in error. Additionally, we believe that the risk of bed-sharing is demonstrated in the British studies as well.

In response to Lawrence's questions on breastfeeding, bed-sharing, and SIDS, we would note that we do not have prevalence data for the Cleveland population. We have evaluated intent to breastfeed at hospital discharge in our data and found no significant difference between those who were bed-sharing and those who were not (20% vs 17%, respectively). We would also point out that the AAP Task Force on Infant Sleep Position and SIDS, in its most recent assessment of factors thought to protect against SIDS, concluded that although there were contradictory reports, the current evidence was insufficient to conclude that breastfeeding was protective for SIDS.⁷

Cindie Carroll-Pankhurst, PhD, MPA
Center for Public Sector Leadership and Service
Mandel School of Applied Social Sciences
Case Western Reserve University
Cleveland, OH 44106-7164

Edward A. Mortimer Jr, MD
Department of Epidemiology and Biostatistics
School of Medicine
Case Western Reserve University
Cleveland, OH 44106-7164

REFERENCES

1. Friedman GD. Primer of Epidemiology. 2nd ed. New York, NY: McGraw-Hill Book Co; 1980:52
2. Dawber TR. The Framingham Study: The Epidemiology of Atherosclerotic Disease. Cambridge, MA: Harvard University Press; 1980
3. Wadsworth MEJ Follow-up of the first national cohort: findings from the Medical Research Council National Survey of Health and Development. Paediatr Perinat Epidemiol 1987; 1:95-117 [Medline]
4. Dingle JH, Badger GF, Jordan WS Jr. Illness in the Home. A Study of 25,000 Illnesses in a Group of Cleveland Families. Cleveland, OH: The Press of Western Reserve University; 1964
5. Blair JS, Fleming PJ, Smith JJ, Babies sleeping with parents: case-control study of factors influencing the risk of sudden death syndrome. BMJ. 1999; 319:1457-1462
6. Leach CEA, Blair JS, Fleming PJ, et al. Epidemiology of SIDS and explained sudden infant deaths. Pediatrics 1999;104(4). Available at: http://www.pediatrics.org/cgi/content/full/104/4/e43
7. American Academy of Pediatrics, Task Force on Infant Sleep Position and Sudden Infant Death Syndrome Changing concepts of sudden infant death syndrome: implications for infant sleeping environment and sleep position. Pediatrics. 2000; 105:650-656 [Abstract/Free Full Text]