A path model of psychosocial and health behaviour change predictors of excessive gestational weight gain

This study aimed to evaluate a conceptual model of psychosocial, behaviour change, and behavioural predictors of excessive gestational weight gain (GWG). Background: Excessive GWG can place women and their babies at risk of poor health outcomes, including obesity. Models of psychosocial and behaviour change predictors of excessive GWG have not been extensively explored; understanding the mechanisms leading to excess GWG will provide crucial evidence towards the development of effective interventions. Method: Two hundred and eighty-eight pregnant women (≤18 weeks gestation) were recruited to a prospective study. Demographic, psychosocial, health behaviour change, and behavioural factors were assessed at 17 (Time 1, T1) and 33 weeks (Time 2, T2) gestation. Pre-pregnancy and final pregnancy weight were obtained and women were classified with/without excessive GWG. Logistic regressions refined the list of predictors of excessive GWG; variables with p < .1 were included in a path analysis. Results: Age, family income, T2 depression, T2 pregnancy-specific coping, T1 buttocks dissatisfaction, T2 GWG-specific self-efficacy, T1 dietary readiness, T1 dietary importance, and T1 vegetable intake predicted excessive GWG in the logistic regressions and were included in the path model. The baseline path model demonstrated poor fit. Once statistically and theoretically plausible paths were added, adequate model fit was achieved (χ² = 21.61(9), p < .05; RMSEA = .07; CFI = .93); this revised model explained 19.5% of the variance in excessive GWG. Women with high T1 buttocks dissatisfaction were more likely to exhibit low levels of dietary readiness. Women with low dietary readiness were more likely to have a lower vegetable intake, which predicted excessive GWG. Women with higher T2 depressive symptoms were more likely to report lower GWG self-efficacy and gain excessively. Conclusion: Future behavioural GWG trials should consider combining psychosocial and health behaviour change factors to optimise GWG.

Small-variance asymptotics for bayesian nonparametric models with constraints

The users often have additional knowledge when Bayesian nonparametric models (BNP) are employed, e.g. for clustering there may be prior knowledge that some of the data instances should be in the same cluster (must-link constraint) or in different clusters (cannot-link constraint), and similarly for topic modeling some words should be grouped together or separately because of an underlying semantic. This can be achieved by imposing appropriate sampling probabilities based on such constraints. However, the traditional inference technique of BNP models via Gibbs sampling is time consuming and is not scalable for large data. Variational approximations are faster but many times they do not offer good solutions. Addressing this we present a small-variance asymptotic analysis of the MAP estimates of BNP models with constraints. We derive the objective function for Dirichlet process mixture model with constraints and devise a simple and efficient K-means type algorithm. We further extend the small-variance analysis to hierarchical BNP models with constraints and devise a similar simple objective function. Experiments on synthetic and real data sets demonstrate the efficiency and effectiveness of our algorithms.