ASSOCIATION BETWEEN INFANT NIGHTTIME-SLEEP LOCATION AND ATTACHMENT SECURITY: NO EASY VERDICT.

We tested whether mother-infant bed-sharing is associated with increased secure infant-mother attachment, a previously unexplored association. Frequency of bed-sharing and mothers' nighttime comforting measures at 2 months were assessed with questionnaires in 550 Caucasian mothers from a population-based cohort. Attachment security was assessed with the Strange Situation Procedure (M.D.S. Ainsworth, M.C. Blehar, E. Waters, & S. Wall, 1978) at 14 months. When using a dichotomous variable, "never bed-sharing" (solitary sleepers) versus "any bed-sharing," the relative risk of being classified as insecurely attached for solitary-sleeping infants (vs. bed-sharers) was 1.21 (95% confidence interval: 1.05-1.40). In multivariate models, solitary sleeping was associated with greater odds of insecure attachment, adjusted odds ratio (OR): 1.50, 95% CI = 1.02-2.20) and, in particular, with greater odds of resistant attachment, adjusted OR = 1.74, 95% CI = 1.10-2.76); and with a lower attachment security score, β = -0.12, t(495) = -2.61, p = .009. However, we found no evidence of a dose-response association between bed-sharing and secure attachment when using a trichotomous bed-sharing variable based on frequency of bed-sharing. Our findings demonstrate some evidence that solitary sleeping is associated with insecure attachment. However, the lack of a dose-response association suggests that further experimental study is necessary before accepting common notions that sharing a bed leads to children who are better or not better adjusted.