Objectives. To describe sleep positions among low birth weight infants, variations in sleep position according to birth weight, and changes in sleep position over time. To analyze risk factors and influences associated with prone sleep.
Design. Prospective cohort study.
Setting. Massachusetts and Ohio, 1995–1998.
Study Participants. Mothers of 907 low birth weight infants.
Results. At 1, 3, and 6 months after hospital discharge, the prevalence of prone sleeping was 15.5%, 26.8%, and 28.3%, respectively. The corresponding rates for supine sleeping were 23.8%, 37.9%, and 50.2% and for side sleeping were 57.3%, 32.4%, and 20.6%. Very low birth weight (VLBW) infants (<1500 g) were most likely to be placed in the prone position. From 1995 through 1998, prone sleeping 1 month after hospital discharge declined among all low birth weight infants from 19.9% to 11.4%; among VLBW infants, the decline in prone sleeping was replaced almost entirely by an increase in side sleeping, whereas in larger low birth weight infants, it was replaced primarily by supine sleeping. Among mothers who placed their infants to sleep in nonprone positions, professional medical advice was cited most frequently as the most influential reason, whereas among mothers of prone-sleeping infants, the infant’s preference was cited most frequently. However, mothers of prone-sleeping VLBW infants also frequently cited the influence of medical professionals and nursery practices as most important in the choice of sleeping position. The factors most strongly associated with prone sleeping were single marital status (odds ratio [OR]: 3.0; 95% confidence interval [CI]: 1.5–6.2), black race (OR: 2.6; 95% CI: 1.5–4.5), birth weight <1500 g (OR: 2.4; 95% CI: 1.3–4.3), and multiparity (OR: 2.1, 95% CI: 1.2–3.5).
Conclusions. Prone sleep decreased among low birth weight infants from 1995 to 1998. However, VLBW infants, who are at very high risk for sudden infant death syndrome, are more likely to sleep prone than larger low birth weight infants.
Sudden infant death syndrome (SIDS) is the leading cause of death among infants beyond the neonatal period. The SIDS-specific infant mortality rate was 66.9 per 100 000 live births in the United States in 1999.1 A number of studies have shown that low birth weight and preterm infants are at especially high risk for SIDS, approximately 3 to 6 times that of full-term, non-low birth weight infants.2–8 Furthermore, the association between prone sleeping and SIDS seems to be even stronger among low birth weight infants than it is known to be among non-low birth weight infants.9 The only published data to date on the sleeping position of low birth weight infants came from the National Infant Sleep Position Study, a telephone survey conducted from 1992 through 1996.10 Data from this study did not demonstrate a difference in the rate of prone sleeping between low birth weight and non-low birth weight infants; however, only 4% of the sample were low birth weight, and additional stratification of the low birth weight sample was not possible.
The American Academy of Pediatrics (AAP) has recommended against prone sleep since 1992. The initial AAP statement regarding positioning and SIDS made a general recommendation that healthy infants be placed on their side or back for sleep.11 The statement, however, mentioned several caveats including that “for premature infants with respiratory distress … prone may well be the position of choice.” In 1996, the AAP modified the previous recommendation to emphasize that supine sleeping was preferred over side sleeping and stated “there are no studies suggesting that recovered preterm infants are exempt from the increased risk of SIDS when placed prone.”12 The AAP strengthened its comments about preterm infants in 2000, stating “There are no data suggesting that strategies designed to reduce risk in full-term infants should not also be applied to premature infants. The relationship to prone sleeping, for example, has been shown to hold for infants of low birth weight as well as for those born with a normal birth weight at term.”13
Indeed, there have been a number of recent studies on the physiologic effects of sleep position in preterm and low birth weight infants that support the AAP’s recommendations. No differences have been observed among preterm infants in the incidence of periodic breathing, apnea, bradycardia, or desaturations below 80% according to sleep position.14,15 Furthermore, preterm and low birth weight infants placed prone have higher heart rates, reduced cardiac variability, and reduced respiratory variability in quiet and active sleep, and less frequent awakenings in quiet sleep compared with those placed supine.14,16,17 These characteristics have been observed in infants who later die of SIDS and may predispose them to a life-threatening event during sleep.18
We used a cohort of infants from the Infant Care Practices Study to describe the prevalence of various sleep positions among low birth weight infants, variations in sleep position according to birth weight, and changes in sleep position trends from 1995 through 1998. We also analyzed risk factors and influences associated with prone sleep.
Infant Care Practices Study is a prospective, longitudinal cohort study with enrollments occurring between 1995 and 1998. A detailed description of the study design has been published previously.19 Briefly, women-infant pairs were recruited during obstetrical admissions in Boston, Lowell, and Lawrence, Massachusetts, and Toledo, Ohio. To obtain the sample of low birth weight infants included in the current study, study personnel reviewed obstetrical records each working day to identify women who had delivered live infants weighing <2500 g in the previous 24 hours. Women were ineligible if they resided out-of-state or intended to move out-of-state in the subsequent 6 months, were not to be caring for the child, were not fluent in English or Spanish, or had a child with a major congenital malformation. Eligible women were contacted according to a list ordered by the terminal 2 digits of the mother’s medical record number. Baseline data were collected at the time of enrollment and follow-up mail questionnaires were administered at 1, 3, and 6 months after the infant’s hospital discharge. Mothers not responding to mailed questionnaires were interviewed by telephone.
In each of the questionnaires, the question related to sleep position was: “Last evening when you put your infant to bed for the night, how did you place him/her?” Answer options included: “Lying on stomach with face down,” “Lying on stomach with face to side,” “Lying on back,” “Lying on side,” “Propped in a sitting position,” “Other (specify),” and don’t know/can’t remember. For analysis, the first 2 options were combined into the category prone sleep and the other categories were maintained. Respondents were also asked if the last 24 hours had been typical for their infant.
In each questionnaire, respondents were asked: “What helped you decide what position to place your infant in to sleep?” Possible answers included “Doctor,” “Nurse or other medical professional,” “Family members or friends,” “Educational material given in hospital, office, or mail,” “Childbirth/prenatal classes,” “Books,’ ”Magazines, or newspapers,“ ”TV or radio,“ ”Followed what I did for other children,“ ”Followed what the nursery did,“ or ”Other (specify)“; multiple answers were allowed. Respondents were then asked: ”Which was the most important in helping you decide?“ with only 1 answer allowed.
Secular trends in sleep position were analyzed by Mantel Extension χ2 test. Risk factors for prone sleep were analyzed using multiple logistic regression. Variables included in the model were: year and site of enrollment; infant’s sex, gender, birth weight, singleton versus multiple birth, and postconceptional age at hospital discharge; mother’s race/ethnicity, age, education, marital status, and parity; and annual household income. For multiple logistic regression analysis, if the value for a given variable was unknown for <5 subjects, those individuals were included in the reference category. If a value for a given variable was unknown for 5 or more subjects, a separate category for “unknown” was used in the model.
A total of 907 mother-infants pairs was enrolled. Characteristics of the study subjects are summarized in Table 1. The mean postconceptional age of the infants was 34.7 weeks (standard deviation [SD] ± 3.4 weeks) at birth, 36.4 weeks (SD ± 2.0 weeks) at the time of hospital discharge, 42.5 weeks (SD ± 3.6 weeks) at the 1-month interview, 52.0 weeks (SD ± 3.9 weeks) at the 3-month interview, and 66.0 weeks (SD ± 5.0 weeks) at the 6-month interview. Follow-up data at 1, 3, and 6 months after hospital discharge were obtained on 744 (82.0%), 720 (79.4%), and 689 (76.0%) subjects, respectively. Response rates were somewhat lower among women with the following characteristics: black, non-Hispanic race (71.3%, 68.0%, 58.0% at 1, 3, and 6 months, respectively), Hispanic ethnicity (77.3%, 68.0%, 52.6%), educated less than high school (68.7%, 63.1%, 51.4%), single (74.1%, 71.2%, 61.2%), and from households with an annual income <$16 000 (68.9%, 66.7%, 56.8%).
Overall, the number of infants placed to sleep in the prone position the evening before the survey at 1, 3, and 6 months after hospital discharge was 115 (15.5%), 193 (26.8%), and 195 (28.3%), respectively. The corresponding figures for supine sleep were 177 (23.8%), 273 (37.9%), and 346 (50.2%) and for side sleep were 426 (57.3%), 233 (32.4%), and 142 (20.6%). Less than 4% of infants slept in sitting or other positions at any time point. When asked if the previous 24 hours had been typical for their infant, 88.6%, 87.9%, and 92.0% responded affirmatively at 1, 3, and 6 months, respectively. Limiting the analysis to those respondents did not substantially change the results. Stratification by birth weight categories demonstrated that very low birth weight (VLBW) infants (<1500 g at birth) were most likely to be placed in the prone position at all 3 time points, especially the first follow-up (P = .004 at 1 month; Fig 1).
Analysis of the secular trends in sleeping position 1 month after hospital discharge showed a declining prevalence of prone sleep and an increasing prevalence of supine sleep (Fig 2). For the enrollment years 1995 through 1998, the prevalence of prone sleep decreased from 19.9% to 11.4% (P for trend = .03), whereas the prevalence of supine sleeping increased from 13.7% to 36.2% (P for trend <.0001). Secular trends in sleep position 1 month after hospital discharge stratified by birth weight are shown in Fig 3. Among VLBW infants, prone sleeping declined from 33.9% in 1995–1996 to 17.5% in 1997–1998, while side sleeping increased and supine sleeping remained unchanged. Among larger low birth weight infants, prone sleeping and side sleeping declined from 1995–1996 to 1997–1998, while supine sleeping increased.
The factors reported by mothers to have the greatest influence on the choice of infant sleeping position are shown in Table 2. Among women whose infants slept in nonprone positions, physicians, and other medical professionals were cited most commonly as the primary influence (50.0% among mothers of VLBW infants and 47.0% among mothers of infants with birth weight 1500–2499 g). Among those women whose infants slept prone, mothers of VLBW infants most frequently cited infant’s preference (32.1%) and physician or other medical professionals (28.6%), whereas mothers of infants with birth weight 1500 to 2499 g mentioned infant’s preference and experience with previous children most frequently (21.4% each).
Multivariate analysis showed that birth weight <1500 g, black race, single marital status, and multiparity were all significant risk factors for prone sleep 1 month after hospital discharge (Table 3). A similar analysis of risk factors for prone sleep at 3 months after hospital discharge (which included the same variables as the 1-month model plus sleep position at 1 month) identified prone sleep at 1 month as the strongest predictor (odds ratio [OR]: 14.8; 95% confidence interval [CI]: 8.3–26.6]). After adjustment for prone sleep at 1 month, the only factors that were significantly associated with prone sleep at 3 months after hospital discharge were enrollment in 1998 compared with 1995 (OR: 0.52; 95% CI: 0.28–0.96) and maternal age 14 to 17 years compared with 25 to 34 years (OR: 4.1; 95% CI: 1.4–11.9).
This study documents the prevalence of various sleeping positions from 1 to 6 months after hospital discharge in a large and diverse cohort of low birth weight infants in the United States. An understanding of the sleep positions of low birth weight infants is particularly important considering the very high risk for SIDS associated with prone sleep in these infants.9
Overall, the sleeping position of these low birth weight infants is quite similar to that described for non-low birth weight infants, with a preponderance of side sleeping 1 month after hospital discharge and an increasing prevalence of both prone and supine sleeping at 3 and 6 months after hospital discharge as side sleeping becomes less common.19 A striking finding of this study, however, is that infants weighing <1500 g at birth were substantially more likely to be placed to sleep in the prone position than infants with a birth weight of 1500 to 2499 g. At 1 month after hospital discharge, VLBW infants were placed to sleep in the prone position almost twice as often as infants of birth weight 1500–2499 g (25.7% vs 13.6%). Because VLBW infants are at an especially high risk for SIDS (3–4 times the risk of non-low birth weight infants in recent US studies), their propensity for prone sleeping may contribute to significant mortality.7,8
Analysis of the trends in sleeping position 1 month after hospital discharge between 1995 and 1998 shows that the prevalence of both prone and side sleeping decreased while supine sleeping increased substantially. Notably, the decline in prone sleep over time was most dramatic in VLBW infants. These trends may reflect the impact of the “Back to Sleep” campaign, as well as the 1996 AAP statement on positioning and SIDS which recommended supine sleep as the preferred position for term and preterm infants alike.12
As prone sleeping declined over time among the VLBW infants, there was a corresponding increase in side sleeping with no change in supine sleeping. In contrast, for infants weighing 1500 to 2499 g at birth, the decline in prone sleeping over time was accompanied by a larger increase in supine sleeping and a decline in side sleeping. Thus, as of 1997–1998, while caregivers of low birth weight infants were generally moving away from the prone sleeping position, those caring for VLBW infants were reluctant to embrace the supine position.
In terms of influences on the choice of sleeping position, mothers who placed their infants in nonprone positions most frequently mentioned physicians and other medical professionals as the most important influence. Among those who placed their infants prone, the infant’s preference was cited most commonly, suggesting that many mothers perceive the prone position as the most comfortable for their infants. However, it is also notable that mothers of prone-sleeping VLBW infants mentioned the recommendation of a physician or other medical professional 28.6% of the time and following the practices of the nursery 10.7% of the time, compared with only 11.9% and 7.1%, respectively, among prone sleeping infants with higher birth weight. One explanation for this disparity is that VLBW infants suffer from a higher prevalence of certain medical conditions, such as gastrointestinal reflux and upper airway problems, that leads some medical professionals to recommend the prone sleep position. It is also possible that physicians, neonatal intensive care nurses, and other medical professionals remain uncomfortable recommending nonprone sleeping for VLBW infants, despite the AAP’s recommendations and the physiologic data that support it.
Among infant characteristics, birth weight <1500 g was a strong predictor of prone sleep one month after hospital discharge. Several maternal characteristics were also associated with prone sleep including black race, single motherhood, and multiparity, all with a two- to threefold higher risk. Maternal age <18 years old was associated with a similar, although not statistically significant, increase in the risk of prone sleep. These maternal risk factors for prone sleep are consistent with those previously identified in non-low birth weight infants.10,19
In analyzing the risk factors for prone sleep among low birth weight infants, we chose to focus on the data collected 1 month after hospital discharge. This time point best represents the time of transition from hospital to home care for these infants and also the point at which interventions may have been implemented. When we examined risk factors for prone sleep at 3 months after hospital discharge, by far the greatest predictor was prone sleep at 1 month after hospital discharge, suggesting that once a sleep position is established for a given low birth weight infant, that choice dwarfs the influence of any other infant or maternal characteristic in determining future sleep position.
The present analysis provides the first detailed evaluation to date of sleep position among low birth weight infants, and it is strengthened by the relatively large size of the cohort and its diversity. There are several limitations, however. Subjects were enrolled only in Massachusetts and Ohio and infant sleep patterns may vary in other parts of the country. The Hispanic participants in the study were mainly of Puerto Rican and Dominican origin, with few Mexican Americans, and the representation of Asian Americans in the cohort was small. Response rates to the survey were good (82.0%, 79.4%, and 76.0% at 1, 3, and 6 months, respectively), but loss to follow-up was proportionately greater among minorities, single women, and those with lower income and less education.
Overall, these data demonstrate that although the prevalence of prone sleeping among low birth weight infants declined from 1995 through 1998, VLBW infants, who are at the highest risk for SIDS, were the most likely to sleep prone. Furthermore, the decline in prone sleeping among VLBW infants over this time period was accompanied by an increase in side sleeping, not supine sleeping. It also seems from these data that physicians and nurses caring for VLBW infants frequently recommend the prone sleeping position. Additional research that quantifies the cumulative risks and benefits—in terms of SIDS as well as aspiration, airway obstruction, apnea, and other medical problems—of various sleep positions among low birth weight (especially VLBW) infants may be needed to help physicians and other medical personnel to make evidence-based recommendations for these at-risk infants.
This work was supported by contract N01-HD-4–3221 from the National Institute of Child Health and Human Development and the National Institute on Deafness and Other Communication Disorders, National Institutes of Health, Rockville, Maryland.
We thank Sandra Hatfield, Dottie Powers, and Debra Zagaeski for research assistance and Maria Francescon, MPH, Patricia Brousseau, Chris DeArmond, Cynthia Nagle, Grace Adeya, and Heather Wightman for recruiting subjects and conducting interviews. We are indebted to the physicians and nurses at the following hospitals: Boston Medical Center and Beth Israel Hospital, Boston, Massachusetts; Lowell General Hospital, Lowell, Massachusetts; Lawrence General Hospital, Lawrence, Massachusetts; and St Vincent’s Medical Center and Toledo Hospital, Toledo, Ohio.
- Received February 27, 2002.
- Accepted July 12, 2002.
- Reprint requests to (L.V.) Slone Epidemiology Center at Boston University, 1010 Commonwealth Ave, Boston, MA 02215. E-mail:
Dr Hunt is currently at National Center on Sleep Disorders Research, National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland.
The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US government.
- ↵Hoyert DL, Freedman MA, Strobino DM, Guyer B. Annual summary of vital statistics: 2000. Pediatrics.2001;108 :1241– 1255
- ↵Malloy MH, Hoffman HJ. Prematurity, sudden infants death syndrome, and age of death. Pediatrics.1995;96 :464– 471
- ↵Malloy MH, Freeman DH. Birth weight- and gestation age-specific sudden infant death syndrome mortality: United States, 1991 versus 1995. Pediatrics.2000;105 :1227– 1231
- ↵Oyen N, Markestad T, Skaerven R, et al. Combined effects of sleeping position and prenatal risk factors in sudden infant death syndrome: the Nordic Epidemiological SIDS Study. Pediatrics.1997;100 :613– 621
- ↵American Academy of Pediatrics, Task Force on Infant Positioning and SIDS. Positioning and SIDS. Pediatrics.1992;89 :1120– 1126
- ↵American Academy of Pediatrics, Task Force on Infant Positioning and SIDS. Positioning and sudden infant death syndrome (SIDS): update. Pediatrics.1996;98 :1216– 1218
- ↵American Academy of Pediatrics, Task Force on Infant Positioning and SIDS. Positioning 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
- ↵Goto K, Mimiran M, Adams MM, et al. More awakenings and heart rate variability during supine sleep in preterm infants. Pediatrics.1999;103 :603– 609
- Copyright © 2003 by the American Academy of Pediatrics