The Contribution of Prone Sleeping Position to the Racial Disparity in Sudden Infant Death Syndrome: The Chicago Infant Mortality Study
Background. Rates of sudden infant death syndrome (SIDS) are over twice as high among African Americans compared with Caucasians. Little is known, however, about the relationship between prone sleeping, other sleep environment factors, and the risk of SIDS in the United States and how differences in risk factors may account for disparities in mortality.
Objective. To assess the contribution of prone sleeping position and other potential risk factors to SIDS risk in a primarily high-risk, urban African American population.
Design, Setting, and Population. Case-control study consisting of 260 infants ages birth to 1 year who died of SIDS between November 1993 and April 1996. The control group consists of an equal number of infants matched on race, age, and birth weight. Prospectively collected data from the death scene investigation and a follow-up home interview for case infants were compared with equivalent questions for living control participants to identify risk factors for SIDS.
Main Outcome Measures. Risk of SIDS related to prone sleeping position adjusting for potential confounding variables and other risk factors for SIDS, and comparisons by race-ethnicity.
Results. Three quarters of the SIDS infants were African American. There was more than a twofold increased risk of SIDS associated with being placed prone for last sleep compared with the nonprone positions (odds ratio [OR]: 2.4; 95% confidence interval [CI]: 1.6–3.7). This OR increased after adjusting for potential confounding variables and other sleep environment factors (OR: 4.0; 95% CI: 1.8–8.8). Differences were found for African Americans compared with others (OR: 1.8; 95% CI: 1.2–2.6 and OR: 10.3, 95% CI: 10.3 [3.2–33.8, respectively]). The population attributable risk was 31%. Fewer case mothers (46%) than control mothers (64%) reported being advised about sleep position in the hospital after delivery. Of those advised, a similar proportion of case mothers as control mothers were incorrectly told or recalled being told to use the prone position, but prone was recommended in a higher proportion of black mothers (cases and controls combined) compared with nonblack mothers.
Conclusions. Prone sleeping was found to be a significant risk factor for SIDS in this primarily African American urban sample, and approximately one third of the SIDS deaths could be attributed to this factor. Greater and more effective educational outreach must be extended to African American families and the health personnel serving them to reduce prone prevalence during sleep, which appears, in part, to contribute to the higher rates of SIDS among African American infants.
Reducing racial disparities in infant mortality is a priority in the United States.1 The leading cause of death in the postneonatal period, sudden infant death syndrome (SIDS), occurs at a rate in African Americans of over twice that in Caucasians.2 This has been true even with the declines in SIDS rates that have occurred among all racial-ethnic groups after introduction of the national “Back to Sleep” campaign.2 This campaign, while including a number of risk reduction messages, has primarily focused on informing the public and health care communities about avoiding the prone position for infant sleep.3 This recommendation was based primarily on research conducted abroad,4 as US studies on the relationship between prone sleeping and SIDS have been limited.5–7
The Chicago Infant Mortality Study was conducted to address this gap in knowledge by determining the contribution of prone sleeping position and other potential risk factors in a primarily high-risk, urban African American population. At the time this study was conceived, rates of SIDS and other causes of postneonatal mortality in Chicago were alarmingly high compared with the national rates (in 1991 the Chicago SIDS rate was 2.6 per thousand live births, compared with the US rate of 1.3).2,8 Furthermore, the disparity in rates between black and white infants was of great concern to the public health community (the 1991 rate for blacks was 4.0 compared with 1.7 for whites),8 and information about how to potentially reduce this disparity was needed. This study describes results related to infant sleeping position and SIDS, including the magnitude of prone sleep as a risk factor for SIDS, comparisons between black and nonblack subjects, and how sleep position choices were made by parents.
DESIGN AND METHODS
Subjects and Data Collected
The National Institute of Child Health and Human Development definition of SIDS was used in this study, ie, “the sudden death of an infant under one year of age, which remains unexplained after a thorough case investigation, including performance of a complete autopsy, examination of the death scene, and review of the clinical history.”9 Infants eligible for inclusion in the study were all Chicago residents who died suddenly and unexpectedly from any cause. To increase the number of SIDS infants for analysis, the age eligibility for infants dying of SIDS was extended to birth, ie, all infants birth to 1 year of age were included. Although no criteria were established a priori for minimum age or to exclude infants who were still in the newborn nursery, the youngest infant dying from SIDS was 5 days old and all died after discharge from the newborn nursery. Infants were enrolled who died during the period November 1993-April 1996 and were handled by the Office of the Medical Examiner of Cook County (Chicago). As required by the Illinois Coroner’s Act of 1977, all sudden unexpected infant deaths must be investigated by the coroner or medical examiner in each respective jurisdiction. In Chicago, therefore, all SIDS and other sudden unexpected deaths are handled by the medical examiner, resulting in 100% case ascertainment.
Protocols for standardized scene investigation, autopsy and medical record review were developed from existing protocols, and modified and implemented for the study.10–22 Medical records obtained included labor and delivery, emergency department, hospital admissions if any, outpatient visits, and ambulance run sheets. The autopsy included gross and microscopic examination, and laboratory and radiographic studies. An extensive death scene investigation with close to 400 questions was conducted as soon as possible after the medical examiner received notification of the infant’s death. It included detailed questions about the events preceding and at discovery of the deceased infant; the medical history of the infant and family; the mother’s prenatal and alcohol, tobacco, and drug use history; and other factors pertinent to determining the cause of death. Using a doll, Polaroid (Cambridge, MA) photographs were taken to indicate the location and position of the infant when found.
Information about the family’s previous contact with the Illinois Department of Children and Family Services relating to abuse or neglect was obtained from that agency. Previous sudden child deaths that occurred in Cook County within the same family were identified by searching the Medical Examiner’s log files against the mother’s last name and last known address.
The cause of death for each infant was determined by the responsible medical examiner with input from the others. One in 4 cases (118/469) were reviewed blindly by an external review team, including a forensic pathologist and a pediatric pathologist experienced in sudden infant death. Diagnoses that differed from those of the medical examiner or each other were discussed by a multidisciplinary committee to establish the final diagnosis. An additional 93 cases were reviewed by the committee to ensure diagnostic consistency; these included infants who were bed sharing during the last sleep, all “undetermined” diagnoses, and other complicated scenarios. There was 89% agreement on the reviewed cases (187 of 211); 9 cases originally diagnosed as SIDS by the Medical Examiner’s Office were changed to undetermined, and 10 diagnosed as other than SIDS were changed to SIDS.
Approximately 2 weeks after the death, a separate interview was conducted with the primary caregiver in the home. This consisted of 235 questions that focused on issues not included in the scene investigation, such as routine sleep habits of the infant, social stressors and supports, and access to health care.
One living control infant was matched to each case infant on, in order of priority: maternal race-ethnicity, age at death/interview, and birth weight within the same category (<2500 g, 2500–<4000 g, and ≥4000 g, and ± 250 g if in the middle category). Potential control infants were identified through ongoing review of birth certificates at the Chicago Department of Public Health. Infants who met the matching criteria were randomly selected in groups of 20 for white infants and in groups of 40 for Hispanic and African American infants, based on experience gained during the pilot phase of the study. The mothers of these infants were contacted by mail and invited to participate. Mothers who responded to our invitation were interviewed on a “first come” basis, and once a control mother for a given case infant was interviewed, any additional control mothers for that infant were notified that they would not be needed. The interviews were timed to take place so that the control infant would ideally be within 2 weeks of the age at death for the case infant, and no greater than 4 weeks. A home interview was conducted using questions taken from the death scene investigation and follow-up interview, appropriately selected, and worded for a living infant. A reference sleep period was identified for the control infant to coincide with the time of day during which the respective case infant was found unresponsive; all questions about “last sleep” including position referred to this reference sleep. Control infants’ mothers received a small stipend for their participation ($20-$35 over the course of the study).
The study was developed and approved by a steering committee consisting of 14 representatives from participating institutions and other experts on infant mortality and urban minority health. The protocol was approved by the institutional review board of the Loyola University Medical Center. The protocols and questionnaires were approved by the National Institute of Child Health and Human Development and the Centers for Disease Control and Prevention. The Chicago Department of Public Health and Office of the Medical Examiner of Cook County approved the components in which they participated. Protocols conducted through the Medical Examiner’s Office (autopsy, death scene investigation, and medical record review) were routine parts of the investigation of sudden infant death and participation did not require informed consent. Informed consent was obtained from case family participants for the follow-up interview and from all control family participants.
Statistical Analysis and Sample Size Calculations
An approximate sample size was determined for the number of SIDS cases to be studied, based on a number of potential risk factors for SIDS. In previous research on sleep position and SIDS in New Zealand, 65% of SIDS cases had slept prone versus 45% of all infants.23 Using the formula for matched case-control studies24 with an α level of 0.05 and a power of 0.90, it was determined that the study would need 178 SIDS cases and 178 matched controls to detect an odds ratio (OR) of 2. The enrollment period for new cases was 30 months based on the Medical Examiner’s previous experience in locating ∼90% of families for death scene investigations, the annual number of SIDS deaths in Chicago, and the need for a larger number of cases to conduct multivariate analysis.
The Cochran Mantel-Haenszel statistic was used to estimate χ2 in comparisons of binary and nominal-scaled variables. Interval data were compared by using independent sample t tests. For both the univariate and multivariate analysis, the matching was taken into account by using conditional logistic regression. Unadjusted and adjusted ORs and corresponding 95% confidence intervals (CIs) were calculated. The population attributable risk was calculated for multiple risk factors using the method of Bruzzi and colleagues.25 All analyses were conducted using SAS/STAT, release 6.12 (SAS Institute, Inc, Cary, NC).
A number of variables potentially confounding the associations between risk factors and SIDS were identified both from the data and previous publications. These factors included: infant’s sex; maternal education; maternal age at her infant’s birth; marital status at the time of interview; parity; adequacy of prenatal care as measured by the Kessner/Institute of Medicine Adequacy of Prenatal Care Index26; job prestige; and median household income (Table 1).11,27–29 Maternal and paternal occupation were classified by the Standard International Occupational Prestige Scale.30 According to the scale, a prestige score of 16 was assigned for nonworking single mothers living from public assistance. For married couples, the higher of maternal or paternal scores was chosen. Estimates of the incomes of the study children’s families were based on the mother’s address at the time of death/interview. Using mapping software (Mapinfo Professional, v.4.5, Mapinfo Corp, Troy, NY), each address was placed within the appropriate block group and the median household income for that block group was abstracted from 1990 census data.31,32
The statistical effects of these 8 potential confounders on the exposure-disease association were studied in detail, both individually and in combination. Using change-in-estimates methods,33–35 a factor was determined to confound the association if the OR was altered by >10% after controlling for that individual factor. Seventeen exposure variables were analyzed: maternal drug, alcohol, or tobacco use in pregnancy and infant factors including breastfeeding, bed sharing, sleep location, environmental tobacco exposure, pillow use, sleep position, soft bedding, pacifier use, recent illness (cold or runny nose, wheezing, cough, or a composite illness score), swaddling, and sweating. For these variables, the ORs adjusted on 4 confounders (maternal education, age, marital status, and Kessner Index of prenatal care) differed <10% from the ORs adjusted on all 8. This reduced set of confounder variables was used in all subsequent logistic models.
Using computerized birth certificate data, analyses were performed that compared several potentially confounding variables between responding and nonresponding control mothers, including weight, maternal race-ethnicity, education, marital status and education, smoking or drinking during pregnancy, high-risk pregnancy, parity, sex of the infant, abnormal birth outcome, and Kessner Index of prenatal care. The first set of analyses compared the mothers who were enrolled (n = 249 for whom comparable computerized birth certificate data were available) with those who responded their willingness to participate but were not needed (n = 321). There were no differences in the distribution of any of the variables between the 2 groups. The second set compared the enrolled responders with those who did not reply to our request (9973). Differences were found in the following variables: maternal education (responders were more highly educated, P = .001), parity (responders had slightly lower parity, P = .027), and Kessner Index (65.2% of responders and 53.6% of nonresponders had adequate care, P = .001).
Forty-seven percent of the control infants’ families were interviewed within 12 weeks of the respective case infant’s death, and 63% were interviewed within 26 weeks. Seventeen percent were interviewed more than a year later. This resulted from a staffing shortage early in the study, during which time greater effort was concentrated on collecting data for the case infants. An analysis of prone sleeping position by the time between the SIDS death and matched control interview revealed no differences in prone rates by lag time for either African American or white case and control infants.
Four hundred sixty-nine infants met the study inclusion criteria and were enrolled in the study. The leading cause of death was SIDS, found for 260 (55%) of the infants. The other causes of death were as follows: undetermined (50, 11%), respiratory tract infections (34, 7%), asphyxia (25, 5%), congenital anomalies (19, 4%), other infections (13, 3%), other injuries (12, 3%), malnutrition attributable to neglect (12, 3%), dehydration (10, 2%), prematurity-related causes (9, 2%), and fire-related causes (7, 1%). (Totals do not add up to 100% because of rounding.) The 260 SIDS infants are the subject of this study.
By definition, all infants who died from SIDS had a full autopsy, scene investigation, and medical record review. Infants who appeared to have died from SIDS but for whom scene investigations could not be completed were designated as undetermined. Ninety-eight percent of autopsies were completed within 1 day of pronouncement of death, and 100% of death scene investigations were completed within 3 days, with the median time to completion 1 day. The 2-week follow-up family interview was conducted for 198 (76%) of the case infants. The median follow-up time for the family interview was 15 days from the date of death. The main reason for not conducting the family interview was inability to locate the primary caregiver (n = 41, 16%), followed by parental refusal (n = 21, 8%). Responders and nonresponders were similar in age, marital status, parity, education, Kessner Index, and infant’s age at death.
The majority of the SIDS infants were African American (75%, n = 195), 13% were white, Hispanic (n = 34), and 12% were white, non-Hispanic (n = 31). There was a perfect match on race-ethnicity for control infants. The age difference between case and control infants was within 2 weeks for 88% of the infant pairs. Only 1 control infant was over the 28-day target (29 days younger than the case infant to whom he was matched). There were no differences in the closeness of age matching by the age at death of the case infants. The mean age of the case infants was 89 days (standard deviation [SD]: 57) and of the control infants 86 (SD: 57, difference not significant). Likewise, matching by birth weight category was very close. Among the case infants, 72 (28%) were <2500 g, 179 (69%) were 2500 to 4000 g, and 9 (3%) were >4000 g. The comparable figures for the control infants were 27%, 70%, and 4%, respectively (difference not significant). The mean birth weights of the cases and controls were 2813 g (SD: 705) and 2919 g (SD: 597), respectively (P = .06).
The majority (88%) of SIDS deaths occurred within 4 months of age, with a peak between 1 to 3 months. Seven percent occurred in the first month, and 2 infants (0.8%) died in the first week (ages 5 and 6 days). The distribution for African American infants is skewed slightly to the younger ages compared with the other infants (Fig 1), but this difference is not statistically significant (P = .67). The mean age of the African American SIDS victims was 12.1 weeks (SD: 7.8) and for the others it was 14.4 weeks (SD: 9.1; P = .048). Differences in gender or gestational age did not explain this difference. However, infants of all race-ethnicities who bed shared were found to be younger than those who did not bed share during the last/reference sleep, and a greater proportion of African American infants bed shared (58.0%) compared with the others (29.2%; P = .001). The typical seasonal pattern of SIDS deaths was found, with a fall-winter predominance (64%).
The relationship between several potentially confounding sociodemographic factors and SIDS is shown in Table 1. The control mothers were slightly older than case mothers (24.8 ± 6.5 years vs 23.2 ± 5.4, respectively; P < .002). The lowest risk category for maternal age was 30 years or older, and the highest risk was found for mothers between 20 and 24 (OR: 2.8; 95% CI: 1.6–4.8). Single marital status was associated with greater than a threefold increased risk (OR: 3.5; 95% CI: 2.1–5.7), and less than a high school education and inadequate prenatal care were each associated with a sixfold increased risk (OR: 5.7; 95% CI: 3.4–9.4 and 6.5, 3.5–12.0, respectively). Other sociodemographic factors found to be associated with increased SIDS risk were higher parity, intermediate or low job prestige score, chronic maternal unemployment, and receiving public assistance (Table 1). Although there was a slight predominance of males among the SIDS victims, the OR (1.3) did not achieve statistical significance (95% CI: 0.9–1.8).
The ORs for prone sleeping did not vary significantly by study year, so all years were combined for subsequent analysis. There was over a twofold increased risk associated with being placed for last sleep in the prone position and SIDS compared with the nonprone positions (OR: 2.4; 95% CI: 1.6–3.7; Table 2). Adjusting for the 4 confounders (maternal education, age, marital status, and adequacy of prenatal care) changed the OR and CI minimally (OR: 2.5; 95% CI: 1.5–4.3). No significantly increased risk associated with the side position over supine position was observed (unadjusted OR: 1.0; 95% CI: 0.6–1.6 and adjusted OR: 1.2; 95% CI: 0.7–2.3). In a multivariate model that included other factors associated with increased or decreased risk of SIDS, the OR for prone sleeping over nonprone sleeping was even higher (OR: 4.0; 95% CI: 1.8–8.8). These factors were soft bedding, maternal smoking during pregnancy, pillow use, bed sharing, upper respiratory infection in the previous 2 days, and pacifier use when put down for the last sleep. There was no interaction observed between maternal smoking and prone sleeping position in the risk of SIDS; namely, prone sleeping was a risk factor for infants regardless of their mother’s smoking status.
The proportion of SIDS infants who had been placed prone for last sleep increased with increasing age; this was not found among the control infants (Table 3). Prone sleeping was not associated with a significantly increased risk among the infants under 1 month of age, but was for the infants 1 to 3 months old and ≥4 months (OR: 2.1; 95% CI: 1.2–3.5 and 4.3, 1.4–13.3, respectively).
The OR for prone sleeping differed significantly according the race-ethnicity of the infant. For African Americans, the unadjusted OR was 1.8 (95% CI: 1.2–2.6), whereas for all others it was 10.3 (3.2–33.8) (Table 4). The respective adjusted ORs were 1.7 (95% CI: 1.1–2.8) and 12.2 (95% CI: 2.0–73.2) (Table 4). Although the rates of prone sleeping for both groups was almost identical among the SIDS infants, 58% among African Americans and 55% among the others, the rates for the control infants were very different, 43% compared with 12%, thus accounting for the different ORs.
The population attributable risk for prone sleeping was 31% based on the multivariate model. That is, 31% of the SIDS cases can be attributed to the risk factor of prone sleeping. The attributable risk was 19% for African Americans and 12% for the others.
Infants usually placed for sleep in the prone position (defined as the usual position placed for sleep in the previous 2 weeks) were at almost twice the risk for SIDS compared with infants sleeping in other positions (unadjusted OR 1.9; 95% CI: 1.2–2.9 and adjusted OR: 1.8; CI: 1.03–3.1; Table 5). No additional increase in SIDS risk was found for infants who were usually placed nonprone and had been placed prone for last sleep compared with those who had been placed prone usually and for last sleep (Table 5). The sample size for these analyses was smaller (n = 198 matched pairs) because there were fewer SIDS mothers participating in the follow-up family interview during which usual sleep position questions were asked.
Mothers were asked during the follow-up interview how they decided what sleep position to use for their infant. This information is available for 193 case mothers and their matched control mothers. A higher proportion of the control mothers (64%) compared with case mothers (46%) reported being advised about sleep position by a doctor or nurse after delivery (P < .001; Table 6). For both groups, the side position was the most commonly advised, followed by the prone position. For cases and controls combined, a greater proportion of African American mothers than non-African American mothers reported being told to use the prone position (25% vs 7%, respectively; P = .01). Over 94% of case and control mothers reported that they had followed the medical advice. For those mothers who were not told what position to use after delivery or did not follow their doctor’s or nurse’s advice, “past experience” was the most important factor influencing their choice. The advice of friends or relatives and the comfort of their infants were also common reasons for both case and control mothers.
In this comprehensive case-control study of SIDS in a major urban center, the majority of SIDS deaths were among African American infants. Demographic and descriptive characteristics of the infants were similar to those found in other populations and settings both within and outside the United States. These include the majority of deaths occurring by 4 months of age, cold season predominance, and several maternal factors including younger age, single marital status, lower education, less adequate prenatal care, higher parity, and lower job prestige.11,27–29,36–40 Although this sample demonstrated an earlier peak than the typical 2- to 4-month peak, there are indications from other reports that the SIDS age distribution may be changing to a higher proportion of younger cases.41 Our finding that the age of death is younger for African Americans compared with Caucasians is consistent with other studies.42,43 Bed sharing was associated with lower age at death for all ethnic groups in this sample, similar to findings from Norway.41 Because this practice was more common among the African Americans, as reported also by others,44 it may account in part for their younger age.
The unadjusted OR of 2.4 for prone sleep, primarily determined by the African American majority, is similar to that of international studies published in 1992 or earlier, when prone sleeping in the general population was highest.45 In this study, higher rates of prone sleeping among the African American control infants persisted as late as 1996, 2 years after the launching of the national “Back to Sleep” campaign. Our finding of higher rates of prone sleeping among African Americans compared with Caucasians is consistent with other studies.46–50 This study additionally provides evidence directly linking prone sleeping to SIDS among African Americans. The population attributable risk for prone sleeping was higher for African American infants than for others, taking other risk factors for SIDS into account. Thus, prone sleeping could be responsible, in part, for the higher prevalence of SIDS observed among African Americans.
In the current study, prone sleeping became a stronger risk factor with increasing age of the infant. It was not associated with a significantly increased risk in the neonatal period, which could be attributable to the small number of deaths in this age group. Consistent with our data, others have reported highest risk for SIDS related to prone sleep during specific age periods, including 10 weeks or more in England51 and 13 to 24 weeks in the Nordic countries.37 This implies that infants may be most vulnerable to SIDS from prone sleeping during a critical period of maturation of physiologic functioning,37,52–54 and emphasizes the need to maintain the recommendation for supine sleeping until at least 6 months of age, and perhaps throughout infancy.55
In contrast to other studies, we did not find that infants who usually sleep nonprone are at higher risk for SIDS when placed prone for last sleep compared with those who usually sleep prone.37,56 This analysis was based on a smaller sample size because the questions about usual sleep habits were asked in the follow-up interview in which participation was reduced, and lack of statistical power could have contributed to our finding. It is unlikely that the results were biased by nonresponse, because responders and nonresponders were similar on a number of demographic and maternal factors (marital status, parity, education, prenatal care index, maternal age, maternal smoking, and age of infant at death) as well as infant care factors (bedding softness, breastfeeding, prone sleeping, and bed sharing). Additional research is needed to better understand the risk of SIDS associated with placing infants prone who are more accustomed to nonprone sleeping.
Side sleeping was not found to be associated with a significantly increased risk of SIDS, similar to findings from a Tasmanian study.57 The OR of 2.3 found in multivariate analysis, however, was similar to other studies showing ORs associated with side sleeping of ∼2.58–61 Because the current study was designed during a time when side sleeping had not yet been identified as a risk factor, the sample size calculations did not take this into account, and thus our sample may have been too small to detect a statistically significant effect.
As in any retrospective study, recall bias may occur if mothers of SIDS infants recall exposures more thoroughly than mothers of unaffected, healthy infants, thus resulting in an apparent association when there is none. However, prospectively collected data on sleep position have confirmed results from previous retrospective studies, indicating that recall bias has not been a major problem in case-control studies of SIDS.60,62 Furthermore, it has been shown that the length of time lapsed between the exposure and the recall has a greater influence on recall accuracy.63 In this study, parents of both SIDS victims and control infants were interviewed about their infant’s sleep position within a short time of the sleep period. Thus, we do not believe that recall bias was a problem in this study.
There were some early delays in interviewing control families in relation to the timing of the matched case infant’s death. This preceded the start of the “Back to Sleep” campaign, and was well before interventions in Chicago were initiated, particularly those targeting minority communities. Therefore, we believe that there was little likelihood that this would have biased the results, further supported by an analysis that revealed no differences in prone rates by the time between the infant death and control interview.
The control selection process presented a number of challenges, particularly in light of the well-known difficulty in recruiting persons of color into research studies,64–70 as well as the study’s requirement to match on race-ethnicity, birth weight, and age. A number of methods to enroll control infants were tested before choosing the method selected as the most feasible and effective. Because the number of potential controls contacted was large compared with those who were actually enrolled, the possibility of nonresponse bias exists. However, the analyses comparing the nonresponders to responders are reassuring, with differences noted in only 3 variables, ie, the responders were more educated, had better prenatal care, and were of lower parity. Appropriate adjustments to outcomes of interest were made to control for these differences.
What can explain the greater use of the prone position among the case infant families? Although the choice of sleep position is likely multifactorial, the medical community seems to have played a significant role. Fewer case mothers than control mothers reported being advised about sleep position in the hospital after the birth of their infants. Side position was most often recommended for both groups, reflecting the recommendations at the time (the recommendation to use the supine position only was made in 1996),71 as well as the probable continued fear among nursing personnel to recommend the supine position because of perceived aspiration risk. This was demonstrated in a survey of Iowa maternity hospitals, in which 90% recommended either the side or supine position for infant sleeping, with fear of aspiration given as the major reason for this policy.72 In the current study, almost all the mothers reported that they followed the advice given, indicating the strong influence of medical personnel. This was similar to findings by Ray et al73 comparing white higher income private patients with African American lower income clinic patients that a majority in both groups followed medical advice given in the hospital, although fewer of the latter mothers had received instructions to use a nonprone position. Likewise, Johnson and colleagues48 found in a survey of inner-city parents of color that medical personnel were influential in promoting the supine sleeping position, although a high proportion of mothers were still resistant. In the current study, African American mothers of both case and control infants were more likely to have used the prone position than other racial-ethnic groups.
Prone sleeping position does seem to explain, in part, the higher SIDS rates among blacks in Chicago, and it is likely that this conclusion can be generalized to the disparities in SIDS rates found elsewhere in the United States. Dispersion of the “back to sleep” message, both directly to parents and caregivers and to medical personnel and others who influence parental infant care choices, is necessary to increase supine sleeping among African American infants. Because contact with hospital personnel after normal deliveries is brief, sleep position and other related recommendations must be delivered consistently and reinforced at each medical encounter. It is hoped that these measures will help reduce the gap found in the rates of SIDS among African American infants compared with other racial-ethnic groups in the United States.
This work was supported by the National Institute of Child Health and Human Development and the National Institute on Deafness and Other Communication Disorders under contract number NO1-HD-3–3188 and the Centers for Disease Control and Prevention and the Association of Teachers of Preventive Medicine under cooperative agreement number U50/CCU300860–06.
We gratefully acknowledge the participation of the Office of the Medical Examiner of Cook County, Chicago, Illinois, and the Chicago Department of Public Health. We also thank Daniel McGee, PhD, for statistical and methodologic consultation and Marian Willinger, PhD, and Solomon Iyasu, MBBS, MPH, for thoughtful review of the manuscript. We extend sincere appreciation to the families who graciously participated in this study.
Members of the Steering Committee, Publication Committee, and Pathology Review Panel: Icy Cade, MD, University of Chicago Pritzker School of Medicine, Chicago, Illinois; Katherine Kaufer Christoffel, MD, MPH, Children’s Memorial Hospital and Northwestern University Medical School, Chicago, Illinois, Richard David, MD, Cook County Hospital, Chicago, Illinois; Edmund Donoghue, MD, Office of the Medical Examiner of Cook County, Chicago, Illinois; Samuel P. Gotoff, MD, Rush-Presbyterian St Luke’s Medical Center, Chicago, Illinois; Fern R. Hauck, MD, MS, Loyola University Chicago Stritch School of Medicine, Maywood, Illinois; Howard Hoffman, MA, National Institute on Deafness and Other Communication Disorders, Bethesda, Maryland; Solomon Iyasu, MBBS, MPH, Centers for Disease Control and Prevention, Atlanta, Georgia; Mitra Kalelkar, MD, Office of the Medical Examiner of Cook County, Chicago, Illinois; Robert Kirschner, MD, University of Chicago Pritzker School of Medicine (formerly with the Office of the Medical Examiner of Cook County), Chicago, Illinois; Loretta Lattyak, Loyola University Medical Center, Maywood, Illinois; Theodoric Manley, PhD, DePaul University, Chicago, Illinois; Patricia Mc Feeley, MD, University of New Mexico School of Medicine, Office of the Medical Investigator, Albuquerque, New Mexico; Cathryn Merrick Moore, CDA, Loyola University Chicago Stritch School of Medicine, Maywood, Illinois; John Paton, MD, Columbia-Michael Reese Hospital, Chicago, Illinois; Tonse N. K. Raju, MD, University of Illinois Medical Center, Chicago, Illinois; Ruth Slaughter, RN, MPH, Chicago Department of Public Health, Chicago, Illinois; Terry Solomon, African American Family Commission, Chicago, Illinois; Marie Valdes-Dapena, MD, University of Miami School of Medicine, Miami, Florida; and Marian Willinger, PhD, National Institute of Child Health and Human Development, Bethesda, Maryland.
- Received August 14, 2000.
- Accepted April 26, 2002.
- Reprint requests to (F.R.H.) University of Virginia Health System, Box 800729, Charlottesville, VA 22908-0729. E-mail:
Dr Hauck is currently with the Department of Family Medicine, University of Virginia Health System, Charlottesville, Virginia, Mr Donovan is currently with Covance, Princeton, New Jersey, and Dr Rowley is currently with the Public Health Sciences Institute, Morehouse College, Atlanta, Georgia.
The content of this publication does not necessarily reflect the views or policies of the US Department of Health and Human Services or the Association of Teachers of Preventive Medicine, nor does mention of trade names, commercial products, or organizations imply endorsement by the US government.
- ↵Addressing health disparities: the NIH program of action. Bethesda, MD: National Institutes of Health; 2001. Available at: http://healthdisparities.nih.gov/whatare.html
- ↵Murphy SL. Deaths: Final Data for 1998. National Vital Statistics Reports 2000. Vol 48. No. 11. Hyattsville, MD: National Center for Health Statistics; 2000. Available at: www.cdc.gov/nchs/releases/00news/finaldeath98.htm
- ↵American Academy of Pediatrics, Task Force on Infant Sleep Position and SIDS. Changing concepts of sudden infant death syndrome: implications for infant sleeping environment and sleep position. Pediatrics.2000;105 :650– 656
- ↵Kattwinkel J, Brooks, J, Keenan ME, Malloy M. Infant sleep position and the sudden infant death syndrome (SIDS) in the United States: joint commentary from the American Academy of Pediatrics and selected agencies of the federal government. Pediatrics.1994;93 :820
- ↵American Academy of Pediatrics, Task Force on Infant Positioning and SIDS. Positioning and SIDS. Pediatrics.1992;89 :1120– 1126
- ↵Chicago Department of Public Health, Epidemiology Program. 1998 Chicago Provisional Infant Mortality Data. Chicago, IL: Chicago Department of Public Health; 1999
- ↵DiMaio DJ, DiMaio VJM. Sudden infant death syndrome. In: DiMaio DJ, DiMaio VJM, eds. Forensic Pathology. New York, NY: Elsevier; 1989:289–297
- Maternal and Child Health Branch. Death Scene and Deputy Coroner Investigation Protocol for Sudden Unexpected Infant Deaths: Instruction Manual. Sacramento, CA: State of California, Department of Health Services, Maternal and Child Health Branch; 1991
- Maternal and Child Health Branch. Autopsy Protocol for Sudden Unexpected Infant Death. Sacramento, CA: State of California, Department of Health Services, Maternal and Child Health Branch; 1991
- Office of the Medical Investigator. SIDS/Childhood. Albuquerque, NM: State of New Mexico, Office of the Medical Investigator
- Task Force for the Study of Non-Accidental Injuries and Child Deaths. Protocol for Determining if an Injury is a Result of Child Abuse or Neglect. Springfield, IL: Illinois Department of Children and Family Services; 1987
- Task Force for the Study of Non-Accidental Injuries and Child Deaths. Protocol for Child Death Autopsies. Springfield, IL: Illinois Department of Children and Family Services; 1987
- Stephens BG. Why the autopsy? In: Harper RM, Hoffman HJ, eds. Sudden Infant Death Syndrome: Risk Factors and Basic Mechanisms. New York, NY: PMA Publishing Corp; 1988:135–141
- Anderson TL, Wells SJ. Data Collection for Child Fatalities: Existing Efforts and Proposed Guidelines. Chicago, IL: American Bar Association; 1991
- Illinois Department of Public Health. Sudden Infant Death Syndrome. Coroner/Medical Examiner’s Manual. Springfield, IL: Illinois Department of Public Health; 1988
- ↵Tonkin S, Hassall I. Infant sleeping position and cot death [letter]. Austr Paediatr J.1989;25 :376– 377
- ↵Schlesselman JJ. Case-Control Studies: Design, Conduct, Analysis. New York, NY: Oxford University Press; 1982:150–152
- ↵Bruzzi P, Green SB, Byar P, et al. Estimating the population attributable risk for multiple risk factors using case-control data. Am J Epidemiol.1985;122 :904– 914
- ↵Kessner DM, Singer J, Kalk CE, Schlesinger ER. Infant Death: An Analysis by Maternal Risk and Health Care. Washington, DC: Institute of Medicine and National Academy of Sciences; 1973
- Kraus JF, Greenland S, Bulterys M. Risk factors for sudden infant death syndrome in the US Collaborative Perinatal Project. Int J Epidemiol.1989;18 :113– 120
- ↵Treiman DJ. Occupational Prestige in Comparative Perspective. New York, NY: Academic Press; 1977
- ↵US Bureau of the Census. 1990 Census of Population. STF3A Illinois [CD ROM]. Washington, DC: Department of Commerce, Economics and Statistics Administration; 1992
- ↵Lavin MR. Understanding the Census. A Guide for Marketers, Planners, Grant Writers and Other Data Users. Kenmore, NY: Epoch Books, Inc; 1996
- ↵Mickey RM, Greenland S. The impact of confounder selection criteria on effect estimation. Am J Epidemiol.1989;129 :125– 137
- Maldonado G, Greenland S. Simulation study of confounder-selection strategies. Am J Epidemiol.1993;138 :923– 936
- ↵Rothman KJ, Greenland S. Modern Epidemiology. 2nd ed. Philadelphia, PA: Lippincott-Raven; 1998
- ↵Øyen N, Markestad T, Skjaerven R, et al. Combined effect of sleeping position and prenatal risk factors in sudden infant death syndrome: the Nordic Epidemiological SIDS Study. Pediatrics.1997;100 :613– 621
- Douglas AS, Allan TM, Helms PJ. Seasonality and the sudden infant death syndrome during 1987–9 and 1991–3 in Australia and Britain. Br Med J.1996;312 :1381– 1383
- ↵Mitchell EA, Tuohy PG, Brunt JM, et al. Risk factors for sudden infant death syndrome following the prevention campaign in New Zealand: a prospective study. Pediatrics.1997;100 :835– 840
- ↵Arnestad M, Andersen M, Vege Å, Rognum TO. Changes in the epidemiological pattern of sudden infant death syndrome in southeast Norway, 1984–1998: implications for future prevention and research. Arch Dis Child.2001;85 :108– 115
- ↵Hoffman HJ, Hunter JC, Ellish DJ, Janerich DT, Goldberg J. Adverse reproductive factors and the Sudden Infant Death Syndrome. In: Harper RM, Hoffman HJ, eds. Sudden Infant Death Syndrome. Risk Factors and Basic Mechanisms. New York, NY: PMA Publishing Corp; 1988:153–175
- ↵Lozoff B, Wolf AW, Davis NS. Cosleeping in urban families with young children in the United States. Pediatrics.1984;74 :171– 182
- ↵Mitchell EA. Sleeping position of infants and the sudden infant death syndrome. Acta Paediatr.1993;389(suppl) :26– 30
- ↵Saraiya M, Serbanescu F, Rochat R, Berg CJ, Iyasu S, Gargiullo PM. Trends and predictors of infant sleep positions in Georgia, 1990–1995. Pediatrics.1998;102(3) . Available at: http://www.pediatrics.org/cgi/content/full/102/3/e33
- ↵Fleming PJ, Gilbert R, Azaz Y, Berry PJ, Rudd PT, Stewart A. Interaction between bedding and sleeping position in the sudden infant death syndrome: a population based case-control study. Br Med J.1990;301 :85– 89
- ↵Davidson-Ward SL, Bautista DB, Keens TG. Hypoxic arousal responses in normal infants. Pediatrics.1992;89 :860– 864
- Fleming PJ, Levin MR, Azaz Y, Wigfield R, Stewart AJ. Interactions between thermoregulation and the control of respiration in infants: possible relationship to sudden infant death. Acta Paediatr.1993;389(suppl) :57– 59
- ↵Scher MS, Guthrie RD, Drieger D, Sun M, Sclabassi R. Maturational aspects of sleep from birth to early childhood. In: Beckerman RC, Brouilette RT, Hunt CE, eds. Respiratory Control Disorders in Infants and Children. Baltimore, MD: Williams and Wilkins; 1992:61–75
- ↵Back to Sleep Campaign Brochure, Questions and Answers for Professionals on Infant Sleeping Position and SIDS. Bethesda, MD: National Institutes of Health; 1998
- Fleming PJ, Blair PS, Bacon C, et al. Environment of infants during sleep and risk of sudden infant death syndrome: results of 1993–95 case-control study for confidential inquiry into stillbirths and deaths in infancy. Br Med J.1996;313 :191– 195
- ↵Mitchell EA, Taylor BJ, Ford RPK, et al. Four modifiable and other major risk factors for cot death: the New Zealand study. J Paediatr Child Health.1992;28(suppl 1) :S3– S8
- ↵Brooke H, Gibson A, Tappin D, Brown H. Case-control study of sudden infant death syndrome in Scotland 1992–5. Br Med J.1997;314 :1516– 1520
- ↵Gibbons LE, Ponsonby A-L, Dwyer T. A comparison of prospective and retrospective responses on sudden infant death syndrome by case and control mothers. Am J Epidemiol.1993;137 :654– 659
- Swanson GM, Ward AJ. Recruiting minorities into clinical trials: toward a participant-friendly system. J Natl Cancer Inst.1995;87 :1747– 1759
- ↵American Academy of Pediatrics, Task Force on Infant Positioning and SIDS. Positioning and sudden infant death syndrome (SIDS): update. Pediatrics.1996;98 :1216– 1218
- ↵Hein JA, Pettit SF. Back to sleep: good advice for parents but not for hospitals? Pediatrics.2001;107 :537– 539
- ↵Ray BJ, Metcalf SC, Franco SM, Mitchell CK. Infant sleep position instruction and parental practice: comparison of a private pediatric office and an inner-city clinic. Pediatrics.1997;99(5) : Available at: http://www.pediatrics.org/cgi/content/full/99/5/e12
- Copyright © 2002 by the American Academy of Pediatrics