OBJECTIVE. To report the duration of breastfeeding among a population of Australian women and to identify factors that are associated with the duration of full breastfeeding to 6 months and any breastfeeding to 12 months. METHODS. Participants were 587 women who were recruited from 2 maternity hospitals in Perth and completed a baseline questionnaire just before or shortly after discharge from the hospital. Women were followed up by telephone interview at 4, 10, 16, 22, 32, 40, and 52 weeks postpartum. Data collected included sociodemographic, biomedical, hospital-related, and psychosocial factors associated with the initiation and the duration of breastfeeding. Cox's proportional hazards model was used to identify factors that were associated with the risk for discontinuing full breastfeeding before 6 months and any breastfeeding before 12 months. RESULTS. At 6 months of age, fewer than one half of infants were receiving any breast milk (45.9%), and only 12% were being fully breastfed. By 12 months, only 19.2% of infants were still receiving any breast milk. Breastfeeding duration was independently, positively associated with maternal infant feeding attitudes and negatively associated with breastfeeding difficulties in the first 4 weeks, maternal smoking, introduction of a pacifier, and early return to work. CONCLUSIONS. Relatively few women achieved the international recommendations for duration of full and overall breastfeeding. Women should receive anticipatory guidance while still in the hospital on how to prevent or manage common breastfeeding difficulties and should be discouraged from introducing a pacifier before 10 weeks, if at all. Improved maternity leave provisions and more flexible working conditions may help women to remain at home with their infants longer and/or to combine successfully breastfeeding with employment outside the home.

Although it is well established that early infant feeding has a major influence on the establishment of the gut microbiota, very little is understood about how the introduction of first solid food influences the colonization process. This study aimed to determine the impact of weaning on the faecal microbiota composition of infants from five European countries (Sweden, Scotland, Germany, Italy and Spain) which have different lifestyle characteristics and infant feeding practices. Faecal samples were collected from 605 infants approximately 4 weeks after the introduction of first solid foods and the results were compared with the same infants before weaning (6 weeks of age) to investigate the association with determining factors such as geographical origin, mode of delivery, previous feeding method and age of weaning. Samples were analysed by fluorescencein situhybridization and flow cytometry using a panel of 10 rRNA targeted group- and species-specific oligonucleotide probes. The genusBifidobacterium(36.5 % average proportion of total detectable bacteria),Clostridium coccoidesgroup (14 %) andBacteroides(13.6 %) were predominant after weaning. Similar to pre-weaning, northern European countries were associated with a higher proportion of bifidobacteria in the infant gut microbiota while higher levels ofBacteroidesand lactobacilli characterized southern European countries. As before weaning, the initial feeding method influenced theClostridium leptumgroup andClostridium difficile+Clostridium perfringensspecies, and bifidobacteria still dominated the faeces of initially breast-fed infants. Formula-fed babies presented significantly higher proportions ofBacteroidesand theC. coccoidesgroup. The mode of birth influenced changes in the proportions of bacteroides and atopobium. Although there were significant differences in the mean weaning age between countries, this was not related to the populations of bifidobacteria or bacteroides. Thus, although the faecal microbiota of infants after first complementary foods was different to that before weaning commenced, many of the initial influences on microbiota composition were still evident.

Count data with excess zeros relative to a Poisson distribution are common in many biomedical applications. A popular approach to the analysis of such data is to use a zero-inflated Poisson (ZIP) regression model. Often, because of the hierarchical study design or the data collection procedure, zero-inflation and lack of independence may occur simultaneously, which render the standard ZIP model inadequate. To account for the preponderance of zero counts and the inherent correlation of observations, a class of multi-level ZIP regression model with random effects is presented. Model fitting is facilitated using an expectation-maximization algorithm, whereas variance components are estimated via residual maximum likelihood estimating equations. A score test for zero-inflation is also presented. The multi-level ZIP model is then generalized to cope with a more complex correlation structure. Application to the analysis of correlated count data from a longitudinal infant feeding study illustrates the usefulness of the approach.