Bayesian compositional regression with microbiome features via variational inference.

The microbiome plays a key role in the health of the human body. Interest often lies in finding features of the microbiome, alongside other covariates, which are associated with a phenotype of interest. One important property of microbiome data, which is often overlooked, is its compositionality as it can only provide information about the relative abundance of its constituting components. Typically, these proportions vary by several orders of magnitude in datasets of high dimensions. To address these challenges we develop a Bayesian hierarchical linear log-contrast model which is estimated by mean field Monte-Carlo co-ordinate ascent variational inference (CAVI-MC) and easily scales to high dimensional data. We use novel priors which account for the large differences in scale and constrained parameter space associated with the compositional covariates. A reversible jump Monte Carlo Markov chain guided by the data through univariate approximations of the variational posterior probability of inclusion, with proposal parameters informed by approximating variational densities via auxiliary parameters, is used to estimate intractable marginal expectations. We demonstrate that our proposed Bayesian method performs favourably against existing frequentist state of the art compositional data analysis methods. We then apply the CAVI-MC to the analysis of real data exploring the relationship of the gut microbiome to body mass index.

Understanding the cumulative risk of maternal prenatal biopsychosocial factors on birth weight: a DynaHEALTH study on two birth cohorts.

BACKGROUND: There are various maternal prenatal biopsychosocial (BPS) predictors of birth weight, making it difficult to quantify their cumulative relationship. METHODS: We studied two birth cohorts: Northern Finland Birth Cohort 1986 (NFBC1986) born in 1985-1986 and the Generation R Study (from the Netherlands) born in 2002-2006. In NFBC1986, we selected variables depicting BPS exposure in association with birth weight and performed factor analysis to derive latent constructs representing the relationship between these variables. In Generation R, the same factors were generated weighted by loadings of NFBC1986. Factor scores from each factor were then allocated into tertiles and added together to calculate a cumulative BPS score. In all cases, we used regression analyses to explore the relationship with birth weight corrected for sex and gestational age and additionally adjusted for other factors. RESULTS: Factor analysis supported a four-factor structure, labelled closely to represent their characteristics as 'Factor1-BMI' (body mass index), 'Factor2-DBP' (diastolic blood pressure), 'Factor3-Socioeconomic-Obstetric-Profile' and 'Factor4-Parental-Lifestyle'. In both cohorts, 'Factor1-BMI' was positively associated with birth weight, whereas other factors showed negative association. 'Factor3-Socioeconomic-Obstetric-Profile' and 'Factor4-Parental-Lifestyle' had the greatest effect size, explaining 30% of the variation in birth weight. Associations of the factors with birth weight were largely driven by 'Factor1-BMI'. Graded decrease in birth weight was observed with increasing cumulative BPS score, jointly evaluating four factors in both cohorts. CONCLUSION: Our study is a proof of concept for maternal prenatal BPS hypothesis, highlighting the components snowball effect on birth weight in two different European birth cohorts.

The criterion validity of willingness to pay methods: A systematic review and meta-analysis of the evidence.

BACKGROUND: The contingent valuation (CV) method is used to estimate the willingness to pay (WTP) for services and products to inform cost benefit analyses (CBA). A long-standing criticism that stated WTP estimates may be poor indicators of actual WTP, calls into question their validity and the use of such estimates for welfare evaluation, especially in the health sector. Available evidence on the validity of CV studies so far is inconclusive. We systematically reviewed the literature to (1) synthesize the evidence on the criterion validity of WTP/willingness to accept (WTA), (2) undertake a meta-analysis, pooling evidence on the extent of variation between stated and actual WTP values and, (3) explore the reasons for the variation. METHODS: Eight electronic databases were searched, along with citations and reference reviews. 50 papers detailing 159 comparisons were identified and reviewed using a standard proforma. Two reviewers each were involved in the paper selection, review and data extraction. Meta-analysis was conducted using random effects models for ratios of means and percentage differences separately. Meta-bias was investigated using funnel plots. RESULTS: Hypothetical WTP was on average 3.2 times greater than actual WTP, with a range of 0.7-11.8 and 5.7 (0.0-13.6) for ratios of means and percentage differences respectively. However, key methodological differences between surveys of hypothetical and actual values were found. In the meta-analysis, high levels of heterogeneity existed. The overall effect size for mean summaries was 1.79 (1.56-2.04) and 2.37 (1.93-2.80) for percent summaries. Regression analyses identified mixed results on the influence of the different experimental protocols on the variation between stated and actual WTP values. Results indicating publication bias did not account for differences in study design. CONCLUSIONS: The evidence on the criterion validity for CV studies is more mixed than authors are representing because substantial differences in study design between hypothetical and actual WTP/WTA surveys are not accounted for.

The effectiveness and cost-effectiveness of a complex community sport intervention to increase physical activity: an interrupted time series design.

OBJECTIVES: An effectiveness and cost-effectiveness analyses of two-staged community sports interventions; taster sports sessions compared with portfolio of community sport sessions. DESIGN: Quasi-experiment using an interrupted time series design. SETTING: Community sports projects delivered by eight lead partners in London Borough of Hounslow, UK. PARTICIPANTS: Inactive people aged 14 plus years (n=246) were recruited between May 2013 and February 2014. INTERVENTIONS: Community sports interventions delivered in two stages, 6-week programme of taster sport sessions (stage 1) and 6-week programme of portfolio of community sporting sessions delivered by trained coaches (stage 2). OUTCOME MEASURES: (a) Change in days with ≥30 min of self-reported vigorous intensity physical activity (PA), moderate intensity PA, walking and sport; and (b) change in subjective well-being and EQ5D5L quality-adjusted life-years (QALYs). METHODS: Interrupted time series analysis evaluated the effectiveness of the two-staged sports programmes. Cost-effectiveness analysis compares stage 2 with stage 1 from a provider's perspective, reporting outcomes of incremental cost per QALY (2015/2016 price year). Uncertainty was assessed using deterministic and probabilistic sensitivity analyses. RESULTS: Compared with stage 1, counterfactual change at 21 days in PA was lower for vigorous (log odds: -0.52; 95% CI -1 to -0.03), moderate PA (-0.50; 95% CI 0.94 to 0.05) and sport(-0.56; 95% CI -1.02 to -0.10). Stage 2 increased walking (0.28; 95% CI 0.3 to 0.52). Effect overtime was similar. Counterfactual change at 21 days in well-being was positive particularly for 'happiness' (0.29; 95% CI 0.06 to 0.51). Stage 2 was more expensive (£101 per participant) but increased QALYs (0.001; 95% CI -0.034 to 0.036). Cost per QALY for stage 2 was £50 000 and has 29% chance of being cost-effective (£30 000 threshold). CONCLUSION: Community-based sport interventions could increase PA among inactive people. Less intensive sports sessions may be more effective and cost-effective.

Grouping Gene Ontology terms to improve the assessment of gene set enrichment in microarray data.

BACKGROUND: Gene Ontology (GO) terms are often used to assess the results of microarray experiments. The most common way to do this is to perform Fisher's exact tests to find GO terms which are over-represented amongst the genes declared to be differentially expressed in the analysis of the microarray experiment. However, due to the high degree of dependence between GO terms, statistical testing is conservative, and interpretation is difficult. RESULTS: We propose testing groups of GO terms rather than individual terms, to increase statistical power, reduce dependence between tests and improve the interpretation of results. We use the publicly available package POSOC to group the terms. Our method finds groups of GO terms significantly over-represented amongst differentially expressed genes which are not found by Fisher's tests on individual GO terms. CONCLUSION: Grouping Gene Ontology terms improves the interpretation of gene set enrichment for microarray data.