arXiv subjects

Anne-Louise Ponsonby

Publications and source records attributed to Anne-Louise Ponsonby.

3 recordsLinked to original sources

arXiv2023

Higher Mediterranean diet score is associated with longer time between relapses in Australian females with multiple sclerosis

A higher Mediterranean diet score has been associated with lower likelihood of multiple sclerosis. However, evidence regarding its association with disease activity and progression is limited. Using data from the AusLong Study, we tested longitudinal associations (over 10 years follow-up) between the alternate Mediterranean diet score (aMED) and aMED-Red (including moderate consumption of unprocessed red meat) and time between relapses and disability measured by Expanded Disability Status Scale (EDSS) (n=132; 27 males, 105 females). We used covariate-adjusted survival analysis for time between relapses, and time series mixed-effects negative binomial regression for EDSS. After adjusting for covariates, both higher aMED (aHR=0.94, 95%CI: 0.90, 0.99, p=0.009) and higher aMED-Red (aHR=0.93, 95%CI: 0.89, 0.97, p=0.001) were associated with significantly longer time between relapses in females. Whether specific dietary components of a Mediterranean diet are important in relation to relapses merits further study.

q-bio.OT↗

arXiv2022

Price Formation in Field Prediction Markets: the Wisdom in the Crowd

Prediction markets are a popular, prominent, and successful structure for a collective intelligence platform. However the exact mechanism by which information known to the participating traders is incorporated into the market price is unknown. Kyle (1985) detailed a model for price formation in continuous auctions with information distributed heterogeneously amongst market participants. This paper demonstrates a novel method derived from the Kyle model applied to data from a field experiment prediction market. The method is able to identify traders whose trades have price impact that adds a significant amount of information to the market price. Traders who are not identified as informed in aggregate have price impact consistent with noise trading. Results are reproduced on other prediction market datasets. Ultimately the results provide strong evidence in favor of the Kyle model in a field market setting, and highlight an under-discussed advantage of prediction markets over alternative group forecasting mechanisms: that the operator of the market does not need to have information on the distribution of information amongst participating traders.

q-fin.PR↗

arXiv2017

Bayesian modelling of lung function data from multiple-breath washout tests

Paediatric respiratory researchers have widely adopted the multiple-breath washout (MBW) test because it allows assessment of lung function in unsedated infants and is well suited to longitudinal studies of lung development and disease. However, a substantial proportion of MBW tests in infants fail current acceptability criteria. We hypothesised that a model-based approach to analysing the data, in place of traditional simple empirical summaries, would enable more efficient use of these tests. We therefore developed a novel statistical model for infant MBW data and applied it to 1,197 tests from 432 individuals from a large birth cohort study. We focus on Bayesian estimation of the lung clearance index (LCI), the most commonly used summary of lung function from MBW tests. Our results show that the model provides an excellent fit to the data and shed further light on statistical properties of the standard empirical approach. Furthermore, the modelling approach enables LCI to be estimated using tests with different degrees of completeness, something not possible with the standard approach. Our model therefore allows previously unused data to be used rather than discarded, as well as routine use of shorter tests without significant loss of precision. Beyond our specific application, our work illustrates a number of important aspects of Bayesian modelling in practice, such as the importance of hierarchical specifications to account for repeated measurements and the value of model checking via posterior predictive distributions.

stat.AP↗