Infant Sleeping Arrangements and Practices During the First Year of Life

METHODS

Data-collection methods for the Infant Feeding Practices Study II (IFPS II) are described in detail elsewhere in this supplement.¹⁷ Questions about infant sleeping arrangements were asked in the postnatal questionnaires administered at 3, 6, 9, and 12 months; data items from these 4 assessments were included in this analysis. On the 3-month postnatal questionnaire, participants were asked to report infant sleeping arrangements for 4 distinct time periods: at infant ages 2 weeks, 1 month, and 2 months, and “now” (ie, at 3 months of age). On the 6-, 9-, and 12-month questionnaires, participants were asked to report arrangements “in the past 4 weeks.” Questionnaires that were returned outside of the predetermined age ranges were excluded to minimize misclassification of responses according to infant age. Acceptable age ranges were 11 to 21 weeks for 3 months, 22 to 33 weeks for 6 months, 34 to 46 weeks for 9 months, and 47 to 62 weeks for 12 months. Fifty-six responses that were out of the acceptable age range were excluded from the analyses.

For certain items on each of the 4 questionnaires, participants were asked to report sleeping arrangements during the night. If night was not the mother's “major time for sleeping,” they were asked to report sleeping arrangements “during your major sleep period.” Data on infant sleeping habits, sleep position, sleep location, and sleeping surface were collected with identical questions on the 3-, 6-, 9-, and 12-month questionnaires covering the 7 specific infant age ranges as described above. Women who reported that they “ever lie down or sleep with [the] baby at night” were asked to report whether they usually slept with the infant or stayed awake until the infant was asleep and then put the infant in a different place; the location and frequency of lying or sleeping with the infant; and whether other people also usually lay down or slept with her and the infant. At 3, 6, 9, and 12 months, all women were also asked why they did or did not bring the infant to bed with them.

Bed-sharing mother-infant dyads were identified on the basis of answers to several questions as described below. All women were asked if they “ever lie down or sleep with your baby at night” in each time frame, as noted earlier. The response choices for this question were “Yes, with the baby in a co-sleeper,” “Yes, in a bed (standard mattress),” “Yes, in a waterbed,” “Yes, on a mattress on the floor,” “Yes, on a couch or other place that is not a bed,” and “No.” Multiple responses were allowed. Contradictory responses (eg, both “yes” and “no” answers) were assigned as missing and not included in the analyses. (The contradictory response rate was 2% at 2 weeks and gradually decreased to 0.3% at 12 months.) All women who responded “no” were considered to be “non–bed sharers,” and they were asked to skip additional questions about sleeping arrangements. The women who answered “yes” were asked, “When you and your baby lay down or slept together, did you usually: ‘Stay with the baby and also sleep’ or ‘Keep awake until the baby was asleep, or finished feeding and then put the baby somewhere else while you slept?’” Only 1 response was allowed. Women who reported that they “usually stay with the baby and also sleep” formed the group defined as bed sharers in this study. Finally, women who reported that they lay down or slept with the infant in a cosleeper were excluded from the bed-sharing group. We excluded this group because we believe that some mothers might consider using cosleepers as being compliant with safe-sleeping recommendations of the AAP Taskforce on Sudden Infant Death Syndrome, which stated: “Cosleepers (infant beds that attach to the mother's bed) provide easy access for the mother to the infant, especially for breastfeeding, but the safety standards for these devices have not yet been established by the Consumer Product Safety Commission.”² As noted above, bed sharing was defined as mother sleeping when lying with the infant on the same bed or other sleeping surfaces (excluding cosleepers) for her nighttime sleep or during the major sleep period.

Cross-sectional data on infant sleeping arrangements were analyzed in frequency tables stratified according to infant age. With the responses to 3-month data, multivariable logistic regression models were used to test the associations between social and behavioral characteristics and specific sleeping arrangements. We chose data from 3-month questionnaires for multivariable regression analyses because 2 to 4 months remains the peak age for SIDS.²

The covariates included in the models were maternal age (<20 or ≥20 years), education (≤12 or >12 years), self-reported race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, Asian, or other), marital status (married or unmarried), poverty level (<185% or ≥185% of the federal poverty level), parity (nulliparous or multiparous), postnatal smoking status (ever or never), and current breastfeeding status (any or none). Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were estimated for the following outcomes in all 3-month respondents: laying the infant down in a nonsupine position; sleeping in a separate room from the infant; and bed sharing. Among bed sharers, we modeled predictors of lying or sleeping with the infant on “a couch or other place that is not a bed” and predictors of lying or sleeping with the infant and other children in the same bed. All the covariates were entered in each regression model simultaneously, and all were retained, with the exception of maternal age in the other children model, because its results were statistically unstable. All analyses were conducted with SAS 9.1 (SAS Institute, Inc, Cary, NC).

RESULTS

The demographic features of this subset of respondents were not significantly different from those of the entire study, which are described elsewhere in this supplement.¹⁷ For this analysis, >2300 participants answered questions concerning infant sleeping arrangements for the 2-week to 3-month time periods; the number decreased to almost 1800 for the 12-month questionnaire.

Nearly 80% of the mothers of young infants chose the recommended supine position for placing their infant for nighttime sleep, whereas a significant minority (>20%) did not follow this recommendation (Fig 1). Noncompliance rates increased with infant age so that by 12 months, 36% were not placing their infant supine. Infant positioning frequencies for daytime naps were similar to those for nighttime sleep at each age (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1

Infant positioning for nighttime sleep from infant age 2 weeks to 12 months: IFPS II.

Cross-sectional frequencies on the duration of infant sleep, sleep location, and infant sleeping surface are presented in Table 1. A majority of infants slept <4 hours at night during the first month of postnatal age. By 3 months, >80% of the infants were sleeping for ≥5 hours each night. During the first 3 months, a majority (85%) of the infants slept in the same room as their mother, and this frequency declined to 29% by 12 months.

Most infants slept in bassinets in the first month, and the greatest percentages were in cribs thereafter. The frequency of sleeping in a cosleeper was very low throughout the first year, ranging from 3% to <1%.

During the first 3 months, 59% to 65% of the mothers lay down or slept with their infant at night. The most common location for the mother to lie down or sleep with the infant was a bed followed by a “couch or other place that is not a bed.” Approximately 1% of the mothers stated that they lay down or slept with their infant on a waterbed.

More than three fourths of the women who ever (in each time period) lay down or slept with their infant usually slept when lying with the infant (Table 1). Among all women, the rates of bed sharing, as defined previously, ranged from 42% at 2 weeks of age to 27% at 12 months of age.

Additional sleeping arrangements among bed-sharing mother-infant pairs are presented in Table 2. Bed-sharing mothers who slept with their infants only part of the night were more likely to do so during the last part of the night or several short periods than during the first part of the night. Among the women who bed shared, approximately half of them slept with their infant 7 nights per week at each infant age. Approximately 30% to 40% of bed-sharing dyads slept together all night, every night of the week.

More than two thirds of the bed-sharing mother-infant pairs also bed shared with the mother's partner. Approximately 5% of the mothers and infants slept with other children at 2 weeks, and this frequency steadily increased to nearly 15% at 12 months (Table 2).

The 3 most common reasons for bringing the infant into bed cited by the mothers were to calm a fussy infant, to help the infant and/or the mother sleep, and to facilitate breastfeeding (Table 3). The frequencies for the first and third reasons were similar at each age for mothers who slept with their infant and those who stayed awake, whereas the second reason was more common among the mothers who slept with their infant. Only 1% of the mothers stated that doctors or nurses advised them to bring the infant to bed. The most frequently cited reasons for not lying down or sleeping with their infant were safety concerns, perceived difficulty in later placing the infant in a crib, and that it is “not commonly done in my family” (Table 4). A small number (10%–15%) noted that doctors or nurses advised them not to bring the infant to bed with them.

Regression models estimating the independent associations between social and behavioral factors and infant sleeping arrangements at the infant age of 3 months are listed in Table 5. Non-Hispanic black mothers were less likely to use the recommended supine infant sleeping position, and nulliparous women were more likely to use the recommended sleeping position. Mothers with more than a high school education, those who were married, and nulliparous mothers were more likely to place the infant to sleep in a different room, independent of other factors. Women at <185% of the federal poverty level, those who breastfed at 3 months, and non-Hispanic black, Hispanic, and Asian women were less likely to place the infant in a different room. Bed sharing was significantly more likely among lower-income women, breastfeeders, smokers, and non-Hispanic black women but less likely among married women. Among bed-sharing women, sleeping with the infant on a couch or other place that was not a bed was less likely for breastfeeding women and non-Hispanic black mothers. Also among bed-sharing mother-infant dyads, sleeping with other children was more likely by those from a race/ethnicity other than non-Hispanic black or white, Hispanic, or Asian.

DISCUSSION

In this study we found that despite the Back to Sleep recommendation to place infants supine for all sleep periods,³ 1 of 4 mothers used nonsupine positions when placing their infants to sleep at 3 months of age. Because SIDS incidence peaks at 2 to 4 months of age, this finding is concerning. Black ethnicity was a predictor of noncompliance with the supine recommendation, as has been shown in other studies.^2,4,18

Our cohort was not nationally representative and, in particular, it underrepresented ethnic minorities and mothers of low socioeconomic status (SES), among whom rates of SIDS are higher.¹⁹ Thus, the high rates of noncompliance with this supine-position recommendation noted in our study (in a cohort with relatively high SES) should be of greater concern, because the actual rates may be even higher in a general US population.^4,5 Furthermore, high rates of noncompliance were found at the peak ages for SIDS incidence. These findings emphasize the need for both general and focused educational interventions to reduce noncompliance with this important recommendation.

We also found that a large percentage of the infants slept in the same room as their mothers, which is recommended by the AAP to lower the risk of SIDS.² However, a substantial number of infants slept in a separate room. Some researchers have shown that sleeping in the parents' room without bed sharing is associated with an approximate one-third lower risk for SIDS compared with sleeping in a separate room.^20–23 Thus, the AAP recommends room sharing with the infant placed in a separate crib or bassinet for sleep.² A majority of the mothers chose either bassinets or cribs for their infant, and despite being a relatively high-SES cohort, <3% of the women used a cosleeper for their infant.

The high rate of bed sharing found in our study, between 42% at 2 weeks and 27% at 12 months, is another practice not in compliance with the AAP recommendations.² The rates were highest during younger infant ages. Sleeping with infants on particularly unsafe sleeping surfaces (ie, couches or waterbeds) was also prevalent (29% at infant age 2 weeks and declining to 17% by 12 months). Of the bed sharers, 5% to 15% also slept with another child or children for nighttime sleep, a practice that has been shown to be an important risk factor for SIDS.²⁴

To our knowledge, this is the first study to document the reasons that mothers choose or not choose to bring their infant to bed with them throughout the first year of life. One of the 3 major reasons for bed sharing was breastfeeding. The facilitative role of bed sharing on breastfeeding has been noted by other studies and expert groups.^6,25 The balancing of such benefits against the potential risks of SIDS or infant suffocation remains the crux of the ongoing debate on bed sharing and breastfeeding. Calming a fussy infant was another important reason for bed sharing; thus, educational interventions should include information about how these practices can be successfully accomplished without sleeping with the infant. One such method is placing the infant's crib or bassinet close to the parents' bed to facilitate breastfeeding and contact with the infant, as recommended by the AAP and others.²

We also found that the reasons for lying down with the infant were similar between those who stayed awake and those who slept with their infant. It is noteworthy that 14% of the women who bed shared felt that the practice was safer than not bringing the infant to bed with them. This study did not ask questions about how or why they held this belief. In a study of primarily black, low-income women, 29% of the mothers believed that having their infant sleep with an adult helps to prevent SIDS.²⁶

We found that a majority of women who chose not to bed share did so for the reason provided by current AAP guidelines, namely, that they considered the practice unsafe for their infant. The pragmatic reasons for not bed sharing included interference with the mother's sleep and perceived difficulty in transitioning the infant's sleeping location to a crib later.

It has been shown that physicians had a strong influence on caregivers in reducing prone placement of infants for sleep¹⁸; no published studies, to our knowledge, have examined the influence of health care providers on mothers' decisions to bed share. We found that only 10% to 15% of the mothers responded that a doctor or nurse had advised them not to take the infant to bed with them. It is possible that the actual frequency for this variable may be higher, because those who answered that they considered the practice to be unsafe might have received such information through their doctor or nurse. However, if this percentage is an accurate reflection of the involvement of physicians or nurses in providing advice, then it is necessary to direct increased educational efforts toward clinicians. Because this is a newer recommendation from the AAP, practitioners may be either less familiar with this recommendation or less comfortable in discussing it with patients, particularly while it remains controversial whether bed sharing is unsafe for all mother-infant pairs or only under certain circumstances.

The strengths of this study include that it used a large cohort, had a high response rate, and had other features noted in the IFPS II methods article.¹⁷ The limitations of the study were that the cohort was not a probability sample of the US population, and the percentage estimates cannot be fully generalized.

Despite these limitations, the findings provide an important template for developing general and focused interventions to reduce noncompliance with current safe-sleeping recommendations. Health providers need to engage in discussions with their patients to better understand the reasons for the choices they are making with regard to sleeping practices and to ensure that they understand the risks and benefits associated with these practices. Research is needed to assess the effects of such interventions on both sleeping practices and breastfeeding outcomes, as well as on reducing the rates of SIDS and other infant deaths that occur as a result of unsafe sleeping arrangements.