Factors associated with prolonged non-nutritive sucking habits in two cohorts of Brazilian children

BACKGROUND Non-nutritive sucking habits (NNSH) are very common during childhood. However, if these habits were maintained for 36 months of age or more, they are considered to be prolonged (PNNSH) and can cause occlusal, physiological and esthetic changes. There is controversy about their prevalence and whether perinatal, social, demographic and health characteristics influence their onset and duration. So, the objectives of this study are to estimate the prevalence of PNNSH and to evaluate perinatal, early life and school age factors associated with their occurrence in children. METHODS A sample of 1,463 children aged 7-11 years born in Ribeirão Preto (RP-1994) and São Luís (SL-1997/98), Brazil, was reevaluated at school age in 2004/05. Birth weight, gestational age and perinatal variables were obtained at birth. Type of feeding, occurrence and duration of finger and pacifier sucking were recorded retrospectively at school age. PNNSH were defined when persisted for 36 months of age or more. Crude and adjusted prevalence ratios (PR) were estimated by Poisson regression (alpha = 5%). RESULTS Prevalence of PNNSH was higher in RP (47.6%) than in SL (20.2%) - (p < 0.001). Perinatal variables were not associated to PNNSH, whilst female sex (PR = 1.27 in RP; PR = 1.47 in SL) and bottle feeding for 24 months or more (PR = 2.24 in RP; PR = 2.49 in SL) were risk factors in both locations. Breast feeding for 12 months or more (PR = 0.53 in RP; PR = 0.31 in SL) was associated with lower prevalence of PNNSH in both places. In SL, children whose mothers lived in consensual union (PR = 1.62) and worked outside the home (PR = 1.51) showed higher prevalence of PNNSH compared to their counterparts. CONCLUSIONS Prevalence of PNNSH was high especially in RP and was not associated with perinatal variables. In both cities there was an association between female sex, shorter breast-feeding duration, longer bottle feeding duration and higher prevalence of PNNSH.