CiteULike: jyuh Brunner

Bayeswatch: an overview of Bayesian statistics.

PC Austin — 2007-09-16T05:04:14-00:00

J Eval Clin Pract, Vol. 8, No. 2. (May 2002), pp. 277-286.

Increasingly, clinical research is evaluated on the quality of its statistical analysis. Traditionally, statistical analyses in clinical research have been carried out from a 'frequentist' perspective. The presence of an alternative paradigm - the Bayesian paradigm - has been relatively unknown in clinical research until recently. There is currently a growing interest in the use of Bayesian statistics in health care research. This is due both to a growing realization of the limitations of frequentist methods and to the ability of Bayesian methods explicitly to incorporate prior expert knowledge and belief into the analyses. This is in contrast to frequentist methods, where prior experience and beliefs tend to be incorporated into the analyses in an ad hoc fashion. This paper outlines the frequentist and Bayesian paradigms. Acute myocardial infarction mortality data are then analysed from both a Bayesian and a frequentist perspective. In some analyses, the two methods are seen to produce comparable results; in others, they produce different results. It is noted that in this example, there are clinically relevant questions that are more easily addressed from a Bayesian perspective. Finally, areas in clinical research where Bayesian ideas are increasingly common are highlighted.

Conserved co-expression for candidate disease gene prioritization

Martin Oti — 2008-04-24T19:51:08-00:00

BMC Bioinformatics, Vol. 9 (23 April 2008), 208.

Dietary advice for reducing cardiovascular risk.

EJ Brunner — 2008-06-13T02:04:15-00:00

Cochrane database of systematic reviews (Online), No. 4. (2007)

BACKGROUND: Changes in population diet are likely to reduce cardiovascular disease and cancer, but the effect of dietary advice is uncertain. OBJECTIVES: To assess the effects of providing dietary advice to achieve sustained dietary changes or improved cardiovascular risk profile among healthy adults. SEARCH STRATEGY: We searched the Cochrane Central Register of Controlled Trials, DARE and HTA databases on The Cochrane Library (Issue 4 2006), MEDLINE (1966 to December 2000, 2004 to November 2006) and EMBASE (1985 to December 2000, 2005 to November 2006). Additional searches were done on CAB Health (1972 to December 1999), CVRCT registry (2000), CCT (2000) and SIGLE (1980 to 2000). Dissertation abstracts and reference lists of articles were checked and researchers were contacted. SELECTION CRITERIA: Randomised studies with no more than 20% loss to follow-up, lasting at least 3 months involving healthy adults comparing dietary advice with no advice or minimal advice. Trials involving children, trials to reduce weight or those involving supplementation were excluded. DATA COLLECTION AND ANALYSIS: Two reviewers independently assessed trial quality and extracted data. Study authors were contacted for additional information. MAIN RESULTS: Thirty-eight trials with 46 intervention arms (comparisons) comparing dietary advice with no advice were included in the review. 17,871 participants/clusters were randomised. Twenty-six of the 38 included trials were conducted in the USA. Dietary advice reduced total serum cholesterol by 0.16 mmol/L (95% CI 0.06 to 0.25) and LDL cholesterol by 0.18 mmol/L (95% CI 0.1 to 0.27) after 3-24 months. Mean HDL cholesterol levels and triglyceride levels were unchanged. Dietary advice reduced blood pressure by 2.07 mmHg systolic (95% CI 0.95 to 3.19) and 1.15 mmHg diastolic (95% CI 0.48 to 1.85) and 24-hour urinary sodium excretion by 44.2 mmol (95% CI 33.6 to 54.7) after 3-36 months. Three trials reported plasma antioxidants where small increases were seen in lutein and beta-cryptoxanthin, but there was heterogeneity in the trial effects. Self-reported dietary intake may be subject to reporting bias, and there was significant heterogeneity in all the following analyses. Compared to no advice, dietary advice increased fruit and vegetable intake by 1.25 servings/day (95% CI 0.7 to 1.81). Dietary fibre intake increased with advice by 5.99 g/day (95% CI 1.12 to 10.86), while total dietary fat as a percentage of total energy intake fell by 4.49 % (95% CI 2.31 to 6.66) with dietary advice and saturated fat intake fell by 2.36 % (95% CI 1.32 to 3.39). AUTHORS' CONCLUSIONS: Dietary advice appears to be effective in bringing about modest beneficial changes in diet and cardiovascular risk factors over approximately 10 months but longer term effects are not known.

Detection of apoptosis in vivo using antibodies against caspase-induced neo-epitopes.

S Jakob — 2008-05-11T14:37:02-00:00

Methods (San Diego, Calif.), Vol. 44, No. 3. (March 2008), pp. 255-261.

Cell death induction by apoptosis is an important process in the maintenance of tissue homeostasis as well as tissue destruction during various pathological processes. Consequently, detection of apoptotic cells in situ represents an important technique to assess the extent and impact of cell death in the respective tissue. While scoring of apoptosis by histological assessment of apoptotic cells is still a widely used method, it is likely biased by sensitivity problems and observed-based variations. The availability of caspase-mediated neo-epitope-specific antibodies offers new tools for the detection of apoptosis in situ. Here, we discuss the use of immunohistochemical detection of cleaved caspase 3 and lamin A for the assessment of apoptotic cells in paraffin-embedded liver tissue. Furthermore, we evaluate the effect of tissue pretreatment and antigen retrieval on the sensitivity of apoptosis detection, background staining and maintenance of tissue morphology.

Phenome connections

Martin Oti — 2008-03-03T15:17:05-00:00

Trends in Genetics, Vol. 24, No. 3. (March 2008), pp. 103-106.

Human phenomics is about to come of age with studies that systematically assess the overlap and relationships among all human genetic diseases. A recent study by Andrey Rzhetsky and colleagues illustrates the power of phenomics by revealing links between conditions that were thought to be distinct, suggesting that they share a genetic basis. Their results imply that the human phenome can be viewed as a landscape of interrelated diseases, reflecting overlapping molecular causation.

A parametric analysis of ordinal quality-of-life data can lead to erroneous results.

E Kahler — 2008-04-14T04:19:19-00:00

Journal of clinical epidemiology, Vol. 61, No. 5. (May 2008), pp. 475-480.

OBJECTIVE: Measurements from health-related quality-of-life (HRQoL) studies, although usually of an ordered categorical nature, are typically treated as continuous variables, allowing the calculation of mean values and the administration of parametric statistics, such as t-tests. We investigated whether parametric, compared to nonparametric, analyses of ordered categorical data may lead to different conclusions. STUDY DESIGN AND SETTING: HRQoL data were obtained from patients with a diagnosis of asthma (n=192) and chronic obstructive pulmonary disease (COPD; n=88) at two time points. The impact of the group factor (asthma vs. COPD) and the time factor (t1 vs. t2) on HRQoL was analyzed with a metric approach (repeated measures ANOVA) and two ordinal approaches (each with a nonparametric repeated measures ANOVA). RESULTS: Using the metric approach, a significant effect of "group" (P=0.0061) and "time" (P=0.0049) on HRQoL was found. The first ordinal approach (ranked total score) still showed a significant effect for "group" (P=0.0033) with a worse HRQoL for patients suffering from COPD. In the second approach (ranks for each HRQoL item and summed ranks), there were no significant effects. CONCLUSION: Applying simple parametric methods to ordered categorical HRQoL scores led to different results from those obtained with nonparametric methods. In these cases, an ordinal approach will prevent inappropriate conclusions.

The modular nature of genetic diseases

Oti — 2006-12-16T21:27:22-00:00

Clinical Genetics, Vol. 71, No. 1. (January 2007), pp. 1-11.

GeneSeeker: extraction and integration of human disease-related information from web-based genetic databases.

MA van Driel — 2007-11-18T07:58:41-00:00

Nucleic Acids Res, Vol. 33, No. Web Server issue. (1 July 2005)

The identification of genes underlying human genetic disorders requires the combination of data related to cytogenetic localization, phenotypes and expression patterns, to generate a list of candidate genes. In the field of human genetics, it is normal to perform this combination analysis by hand. We report on GeneSeeker (http://www.cmbi.ru.nl/GeneSeeker/), a web server that gathers and combines data from a series of databases. All database searches are performed via the web interfaces provided with the original databases, guaranteeing that the most recent data are queried, and obviating data warehousing. GeneSeeker makes the same selection of candidate genes as the human geneticists would have performed, and thus reducing the time-consuming process to a few minutes. GeneSeeker is particularly well suited for syndromes in which the disease gene displays altered expression patterns in the affected tissue(s).

A text-mining analysis of the human phenome

Marc van Driel — 2006-02-22T15:53:51-00:00

European Journal of Human Genetics, Vol. aop, No. current.

Inflation of the type I error rate when a continuous confounding variable is categorized in logistic regression analyses.

PC Austin — 2007-08-02T05:39:40-00:00

Stat Med, Vol. 23, No. 7. (15 April 2004), pp. 1159-1178.

This paper demonstrates an inflation of the type I error rate that occurs when testing the statistical significance of a continuous risk factor after adjusting for a correlated continuous confounding variable that has been divided into a categorical variable. We used Monte Carlo simulation methods to assess the inflation of the type I error rate when testing the statistical significance of a risk factor after adjusting for a continuous confounding variable that has been divided into categories. We found that the inflation of the type I error rate increases with increasing sample size, as the correlation between the risk factor and the confounding variable increases, and with a decrease in the number of categories into which the confounder is divided. Even when the confounder is divided in a five-level categorical variable, the inflation of the type I error rate remained high when both the sample size and the correlation between the risk factor and the confounder were high.