Secular changes in sleep position during infancy: 1995-1998

Prone sleeping among infants has been associated with an increased risk of sudden infant death syndrome. The objective of this study was to compare factors associated with sleep position in 1995-1996 and 1997-1998 and to assess secular trends in use of prone infant sleep position from 1995 through 1998 among families stratified by race and education.

A prospective cohort study was conducted in eastern Massachusetts and northwest Ohio of 12 029 mothers of infants who weighed > or =2500 g at birth. Descriptive statistics and multivariate odds ratios were used to relate maternal and infant characteristics to prone and supine sleeping.

A total of 14 206 mothers (25% of those eligible) were enrolled. A total of 12 029 mothers (85% of enrolled) responded to the 1-month and 11 552 mothers (81% of enrolled) responded to the 3-month follow-up questionnaire. A decline in use of the prone sleep position and increase in use of the supine position was observed during the 4 years of the study. Factors associated with prone and supine sleep position were similar in 1995-1996 and 1997-1998. In 1997-1998, use of prone sleeping at 1 month of age reached the goal of < or =10% only among infants of white and Asian women, married women, women who were older than 25 years, women who were college graduates, and women with incomes >$55 000 per year. At 3 months of age, however, prone sleeping increased to 12% to 17% in these groups. These same groups were most likely to use the supine position; 38% to 45% were supine at 1 month, increasing to 56% to 64% by 3 months of age. However, as of the end of 1998, approximately 27% of infants of non-college-educated black and Hispanic mothers were placed to sleep in the prone position and only 20% to 30% were being placed to sleep in the supine position at 3 months of age.

Recommendations to avoid prone sleep position and especially the recommendation that supine sleep position is preferred have not been effectively delivered to black and Hispanic families and to families of low-income and less than a college education.