Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Statistical Power”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,009 records · Page 56Linked to original sources
PubMed2007

MTHFR C677T and colorectal cancer risk: A meta-analysis of 25 populations.

The common functional methylenetetrahydrofolate reductase (MTHFR) C677T polymorphism may influence the risk of colorectal cancer (CRC), but data from published studies with individually low statistical power are conflicting. To clarify the role of MTHFR C677T genotype in CRC, we considered all available studies in a meta-analysis. Studies reporting on MTHFR C677T genotype and CRC were searched in PubMed up to April 2006. The principle prior hypothesis was that homozygosity for MTHFR 677TT would be associated with reduced risk of CRC. Data were available for 29,931 subjects, including 12,243 with CRC, from 25 independent populations. Compared to the homozygous CC genotype, the MTHFR 677TT genotype was associated with a reduced risk of CRC (odds ratio (OR): 0.83; 95% confidence interval (CI): 0.75-0.93; p = 0.001). There was some heterogeneity among the results of individual studies, but this was not statistically significant (heterogeneity p = 0.12; I2 = 25.8%). Heterozygosity for MTHFR 677 did not influence CRC risk (OR: 0.99; 95% CI: 0.94-1.04). These findings indicate that individuals homozygous for the MTHFR 677TT genotype are at moderately reduced risk of CRC, and support the proposal that common genetic variation in the MTHFR gene contributes to CRC susceptibility, probably accounting for at least 9% of the total incidence.

Adult↗
PubMed2004

The analysis of repeated 'direct' measures of change illustrated with an application in longitudinal imaging.

The use of repeated measures of an outcome variable to improve statistical power and precision in randomized clinical trials and cohort studies is well documented. Linear mixed models have great utility in the analysis of such studies in many medical applications including imaging. However, in imaging studies and other applications the basic outcome can be a 'direct' measure of change in a variable, as opposed to a difference calculated by subtraction of one measured value from another. The correlation structure of such repeated measures of 'direct' change, in particular the non-independence of within-person consecutive measures, adds complexity to the analysis. In this paper, we present a family of hierarchical mixed models for the analysis of such data and explain how to implement them using standard statistical software. We illustrate the use of our models with data from a cohort of patients with Alzheimer's disease.

Alzheimer Disease↗
PubMed2004

Asymptotic properties of the Fisher kernel.

This letter analyzes the Fisher kernel from a statistical point of view. The Fisher kernel is a particularly interesting method for constructing a model of the posterior probability that makes intelligent use of unlabeled data (i.e., of the underlying data density). It is important to analyze and ultimately understand the statistical properties of the Fisher kernel. To this end, we first establish sufficient conditions that the constructed posterior model is realizable (i.e., it contains the true distribution). Realizability immediately leads to consistency results. Subsequently, we focus on an asymptotic analysis of the generalization error, which elucidates the learning curves of the Fisher kernel and how unlabeled data contribute to learning. We also point out that the squared or log loss is theoretically preferable-because both yield consistent estimators-to other losses such as the exponential loss, when a linear classifier is used together with the Fisher kernel. Therefore, this letter underlines that the Fisher kernel should be viewed not as a heuristics but as a powerful statistical tool with well-controlled statistical properties.

Algorithms↗
PubMed1998

Underemployment and alcohol misuse in the National Longitudinal Survey of Youth.

OBJECTIVE: This study measured the impact of unemployment and underemployment on alcohol misuse. METHOD: A panel of respondents from the National Longitudinal Survey of Youth was studied in 1984-85 and 1988-89. In each pair of years, the effect of employment change (e.g., becoming underemployed) on alcohol misuse was assessed controlling for misuse in the first year. Alcohol misuse was operationalized in two ways: elevated symptoms and heavy drinking. Three samples were analyzed: a core sample of 2,441 who were available in both pairs of years (approximately 65% male) and two extended samples that included everyone available in one pair of years but not the other (n = 4,183 in 1984-85 and n = 3,926 in 1988-89). RESULTS: The 1984-85 analyses revealed a significant association of adverse change in employment with both elevated alcohol symptoms and heavy drinking (the latter moderated by prior heavy drinking). The 1988-89 analyses found no relationship between adverse change in employment and heavy drinking in the core sample and no main effect of adverse change in employment on symptoms, but they did reveal interactions involving prior symptoms (core sample) and marital status (extended sample). CONCLUSIONS: Several explanations for these decreasing effects over time were considered including changes in measurement reliability, statistical power, economic environment and respondents' maturity. These results confirm previous findings that job loss can increase the risk of alcohol misuse, provide new evidence that two types of underemployment (involuntary part-time and poverty-level wage) can also increase this risk and suggest that these effects vary over time.

Adolescent↗
PubMed2002

An introduction to statistical concepts for the analysis of EEG data and the planning of pharmaco-EEG trials.

The use of descriptive statistics and graphics for electroencephalogram (EEG) data after initial data reduction and the application of hypothesis tests to EEG data are described. In planning pharmaco-EEG trials, some notion of statistical power and variability of effect size is necessary as they are used to calculate sample size in parallel-group clinical trials. The need for sensitivity analysis during the study planning phase is demonstrated as well. Because EEG data are highly dimensional data, an outline of methods for reducing dimensionality is provided. These concepts are demonstrated using data from patients with Alzheimer's disease and from healthy controls.

Alzheimer Disease↗
PubMed1989

Results and evaluations of 48 continuous breeding reproduction studies conducted in mice.

Results of 48 continuous breeding reproduction (RACB) studies are summarized and control data from these studies are used to determine the statistical sensitivity of each endpoint from different parts of these studies. Results of testing individual chemicals compared well with results of multigeneration studies reported in the literature. Continuous breeding studies were able to discriminate reproductive toxicants from nontoxicants, and provided valuable structure-activity information. When mice in continuous cohabitation produce multiple litters, the statistical sensitivity of fertility endpoints is quite high and is comparable to that associated with other sensitive indicators of reproductive function, such as testis weight and sperm parameter measures. The principal advantages of the RACB protocol in comparison to multigeneration studies are: (1) the increased sensitivity and statistical power, (2) the ability to monitor progression of toxicity and to detect subfertility, (3) use of a battery of endpoints including sperm measures, (4) the ability to determine the affected sex(es), and (5) slightly reduced testing time.

Animals↗
PubMed2003

Sample sizes for comparing means of two lifetime distributions with type II censored data: application in an aging intervention study.

Sample size determination is a very important part of planning for clinical trials. Most clinical trials do not follow all their subjects to the terminal event, resulting in censored observations. This article presents a method of computing sample sizes required to achieve adequate statistical power to compare the means of two lifetime distributions when both samples are subject to type II censoring. Our approach is based on the location-scale family of log-transformed lifetime distributions as compared to that based on the log-rank test and the family of proportional hazards. Specific applications to log-normal distribution and Weibull distribution are also discussed.

Aging↗
PubMed2001

Assessing gene significance from cDNA microarray expression data via mixed models.

The determination of a list of differentially expressed genes is a basic objective in many cDNA microarray experiments. We present a statistical approach that allows direct control over the percentage of false positives in such a list and, under certain reasonable assumptions, improves on existing methods with respect to the percentage of false negatives. The method accommodates a wide variety of experimental designs and can simultaneously assess significant differences between multiple types of biological samples. Two interconnected mixed linear models are central to the method and provide a flexible means to properly account for variability both across and within genes. The mixed model also provides a convenient framework for evaluating the statistical power of any particular experimental design and thus enables a researcher to a priori select an appropriate number of replicates. We also suggest some basic graphics for visualizing lists of significant genes. Analyses of published experiments studying human cancer and yeast cells illustrate the results.

Computational Biology↗
PubMed1997

Effect of measurement error on energy-adjustment models in nutritional epidemiology.

The use and interpretation of energy-adjustment regression models in nutritional epidemiology has been vigorously debated recently. There has been little discussion, however, regarding the effect of dietary measurement error on the performance of such models. Contrary to conventional assumptions invoked in the standard treatment of the effect of measurement error in regression analysis, reporting errors in dietary studies are usually biased, correlated with true nutrient intakes and with each other, heteroscedastic, and nonnormally distributed. Methods developed in this paper allow for this more complex error structure and are therefore more appropriate for dietary data. For practical illustration, these methods are applied to data from the Women's Health Trial Vanguard Study. The results demonstrate considerable shrinkage in the magnitude of the estimated main exposure effect in energy-adjustment models due to attenuation of the true effect and contamination from the effect of an adjusting covariate. In most cases, this shrinkage causes a sharply reduced statistical power of the corresponding significance test in comparison with measurement without error. These results emphasize the need to understand the measurement error properties of dietary instruments through validation/calibration studies and, where possible, to correct for the impact of measurement error when applying energy-adjustment models.

Aged↗
PubMed1975

A consideration of the chi-square test of Hardy--Weinberg equilibrium in a non-multinomial situation.

The multinomial Dirichlet distribution was used to study the effect of correlation between observations in a sample on the frequency of rejection of the Hardy--Weinberg Law when in fact it was true in the population being sampled. It was shown that the usual X2 goodness-of-fit test of Hardy--Weinberg is very sensitive to non-multinomial sampling. In view of the lack of statistical power of the test to detect deviations due to in breeding, it is likely that whenever H--W is rejected using samples of size 100 or less, the underlying causation is sample correlation rather than failure of the H--W law to be true. Related to these findings is, of course, the effect of pooling heterogeneous frequencies or, in the case of contingency tables, Simpson's paradox (see Simpson, 1951).

Gene Frequency↗
PubMed2006

Carbamazepine for prevention of oxaliplatin-related neurotoxicity in patients with advanced colorectal cancer: final results of a randomised, controlled, multicenter phase II study.

BACKGROUND: Oxaliplatin-induced neurotoxicity is a growing, relevant clinical problem. In this study we evaluated the efficacy and safety of carbamazepine for prevention of oxaliplatin-associated neuropathy in patients with advanced colorectal cancer. METHODS: Chemotherapeutic treatment consisted of oxaliplatin 85 mg/m(2) given biweekly and weekly folinic acid 500 mg/m(2) followed by a 24-h infusion of 5-FU 2000 mg/m(2) (FUFOX). One cycle consisted of six consecutive weeks of treatment followed by two weeks of rest (=Treatment B). For Treatment A carbamazepine was added in a dosage for targeted plasma levels of 4-6 mg/L. Neurotoxicity was regularly assessed using a specific scale. Moreover, an evaluation of chronic sensory symptoms and a neurologic examination including tests for vibrational sense, strength and deep tendon reflexes were added creating a peripheral neuropathy (PNP) score. RESULTS: The prospectively defined adequate number of patients needed to provide power for the primary outcome could not be achieved. 19 patients were assigned to Treatment A and 17 to Treatment B. At baseline, the distribution of all clinicopathologic variables was comparable between the two groups. Overall response rates were 16% and 24% and overall survival 15.1 months and 17.4 months for Treatment A and Treatment B, respectively. Between Treatment A and Treatment B there were no major differences when considering worst neurotoxicity during the study period (p=0.46). Grade 3/4 neurotoxicity occured in 4 patients with Treatment A vs. 6 patients with Treatment B. There were no major differences between both groups in each category of the PNP score. CONCLUSIONS: Based on the small number of patients and low statistical power of our study definite conclusions regarding efficacy and safety of carbamazepine for prevention of oxaliplatin-associated neuropathy in patients with advanced colorectal cancer cannot be drawn.

Adult↗
PubMed1998

Functional magnetic resonance imaging of the human brain: data acquisition and analysis.

It is now feasible to create spatial maps of activity in the human brain completely non-invasively using magnetic resonance imaging. Magnetic resonance imaging (MRI) images in which the spin magnetization is refocussed by gradient switching are sensitive to local changes in magnetic susceptibility, which can occur when the oxygenation state of blood changes. Cortical neural activity causes increases in blood flow, which usually result in changes in blood oxygenation. Hence changes of image intensity can be observed, given rise to the so-called Blood Oxygenation Level Dependent (BOLD) contrast technique. Use of echo-planar imaging methods (EPI) allows the monitoring over the entire brain of such changes in real time. A temporal resolution of 1-3 s, and a spatial resolution of 2 mm in-plane, can thus be obtained. Generally in a brain mapping experiment hundred of brain image volumes are acquired at repeat times of 1-6 s, while brain tasks are performed. The data are transformed into statistical maps of image difference, using the technique known as statistical parametric mapping (SPM). This method, based on robust multilinear regression techniques, has become the method of reference for analysis of positron emission tomography (PET) image data. The special characteristics of functional MRI data require some modification of SPM algorithms and strategies, and the MRI data must be gaussianized in time and space to conform to the assumptions of the statistics of Gaussian random fields. The steps of analysis comprise: removal of head movement effects, spatial smoothing, and statistical interference, which includes temporal smoothing and removal by fitting of temporal variations slower than the experimental paradigm. By these means, activation maps can be generated with great flexibility and statistical power, giving probability estimates for activated brain regions based on intensity or spatial extent, or both combined. Recent studies have shown that patterns of activation obtained in human brain for a given stimulus are independent of the order and spatial orientation with which MRI images are acquired, and hence that inflow effects are not important for EPI data with a TR much longer than T1.

Artifacts↗
PubMed2000

A statistically derived index for classifying East Coast fever reactions in cattle challenged with Theileria parva under experimental conditions.

A statistically derived disease reaction index based on parasitological, clinical and haematological measurements observed in 309 5 to 8-month-old Boran cattle following laboratory challenge with Theileria parva is described. Principal component analysis was applied to 13 measures including first appearance of schizonts, first appearance of piroplasms and first occurrence of pyrexia, together with the duration and severity of these symptoms, and white blood cell count. The first principal component, which was based on approximately equal contributions of the 13 variables, provided the definition for the disease reaction index, defined on a scale of 0-10. As well as providing a more objective measure of the severity of the reaction, the continuous nature of the index score enables more powerful statistical analysis of the data compared with that which has been previously possible through clinically derived categories of non-, mild, moderate and severe reactions.

Animals↗
PubMed2002

Selecting the best method to evaluate bronchodilation when analysing bronchodilator studies.

Changes in lung function (that is, FEV(1)) are commonly expressed as a percentage of the predicted value. The question is whether this parameter is best suited for statistical analysis. A total of 2199 bronchodilation tests were selected and pre- and post-dilator FEV(1), age, height and sex extracted. Predicted FEV(1) values, the difference between pre- and post-dilator FEV(1) and the change as percentage of predicted or baseline FEV(1) were calculated. The latter parameters were subjected to analysis of variance and pre- and post-dilator FEV(1) to analysis of covariance. The statistical power of these four approaches and the residual structure were used to select the best approach. Analysis of variance of the change as percentage of predicted and covariance analysis of pre- and post-dilator FEV(1) offers the best performance. The residuals of the analysis of variance showed a funnel shaped structure. The latter causes residual heterogeneity, resulting in a loss of power and deviations between the actual and nominal alpha-level. This is not taken into account when calculating sample sizes using the standard equations, which implicitly assume lack of these problems. Analysis of covariance does not suffer from these problems. The best way to evaluate bronchodilation in trials is to use analysis of covariance.

Age Factors↗
PubMed2005

MRI quantification in vivo of corticosteroid induced thymus involution in mice: correlation with ex vivo measurements.

Thymus involution is a useful marker of transactivation-mediated side effects in preclinical therapeutic index testing of new anti-inflammatory glucocorticosteroids, and is usually measured post mortem. We have validated the use of MRI for non-invasive in vivo measurement of mouse thymus involution induced by dexamethasone (DEX). Tl-weighted spin echo 7 T images provided satisfactory contrast between thymus and surrounding connective tissue and fat. Increasing doses of DEX caused thymus involution, reflected in MRI volume (87+/-14, 33+/-10, 28+/-6, 16+/-7 microl in dosage groups of Cremophor vehicle, 1, 10 and 30 mg/kg subcutaneous respectively, n=6/group, mean+/-standard deviation) and post mortem wet weight (64+/-12, 33+/-6, 25+/-9, 23+/-8 mg). Correlation between MRI volumes and wet weights was very good (r=0.842). Measuring pre-dose MRI volumes and then assessing DEX effects as post-dose change from baseline produced no statistical advantage relative to considering post-dose MRI thymus volume alone, probably due to variability in pre-dose baseline values compounding post-dose variability. Smaller group sizes were sufficient to achieve a given statistical power using MRI post-dose volume than using wet weight, suggesting a role for MRI in differentiating the effects of compounds which produce similar effects, or in contexts where the use of large groups of animals is impractical or ethically unacceptable.

Animals↗
PubMed2006

A method for identification of the expression mode and mapping of QTL underlying embryo-specific characters.

Embryos of crop seeds are one of the major sources of the plant protein and lipid for human nutrition. The genetic expression for embryo-specific characters in crop seeds can be controlled exclusively by the embryo or the maternal genotypes and sometimes by both simultaneously. However, current methods for mapping quantitative trait loci (QTLs) underlying characters of maternal plants have not been effective in dealing with the QTL analysis of embryo characters. On the basis of the expression feature of embryo, a statistical method was proposed for the identification of expression mode and mapping of QTL controlling embryo traits. The maximum likelihood method implemented via the expectation maximization algorithm was used to estimate parameters of a putative embryo-specific QTL. The QTL expression mode was identified by the likelihood ratio test statistic. Statistical power and other properties of the proposed method were investigated under a variety of scenarios through simulation studies. The results showed that the mapping method neglecting the effects of embryo genotype or maternal effects could neither identify the expression mode of QTL nor estimate its genetic effects accurately, whereas the proposed method could effectively map the embryo-specific QTL of various expression modes.

Algorithms↗
PubMed1991

Hostility and coronary artery disease.

Studies of the association between type A behavior and coronary heart disease have yielded inconsistent findings. A possible explanation for these inconsistent findings is that type A behavior is simply a marker for other behaviors that are truly related to coronary heart disease. Hostility is one such behavior that has been found in several recent studies to predict coronary heart disease and coronary atherosclerosis; however, several other studies have found null results. In the present study, the predictive power of hostility was tested in a study population of hospitalized men (n = 118) and women (n = 40) scheduled for coronary angiography. Potential coronary risk behaviors were assessed in the angiography patients and they were given the type A Structured Interview. Hostility was measured with the Cook-Medley Hostility Inventory and the Behavior Pattern Hostility Index, a measure of hostility derived from the type A Structured Interview. No significant positive associations were found for either Cook-Medley hostility or behavior pattern hostility and coronary occlusion. This was true whether hostility or coronary occlusion was treated as a dichotomous variable or as a continuous variable. In fact, most of the observed associations were opposite to the predicted direction, although none was statistically significant. Replicating cutpoints of the Cook-Medley Hostility Inventory used in other studies that have reported positive associations with coronary heart disease also yielded null findings. The association between hostility and coronary occlusion was slightly modified by age and sex, but the interaction coefficients were not significant. The sample size yielded adequate statistical power to detect the hypothesized associations, and there was no evidence that selection bias, measurement error, or unexamined confounding accounted for the null findings. These results failed to confirm some earlier reports showing a positive association between hostility and coronary artery disease.

Age Factors↗
PubMedDate not supplied

Accumulating evidence from independent studies: what we can win and what we can lose.

When asking 'what is known' about a drug or therapy or program at any time, both researchers and practitioners often confront more than a single study. Facing a variety of findings, where conflicts may outweigh agreement, how can a reviewer constructively approach the task? In this discussion, I will outline some questions that can only be answered by examining a group of independent studies. I will also discuss some pitfalls that sometimes swamp the benefits we can gain from synthesis. Most of these pitfalls are avoidable if anticipated early in a review. The benefits of a quantitative review include information about how to match a treatment with the most promising recipients; increasing the statistical power to detect a significant new treatment; telling us when 'contextual effects' are important; helping us to assess the stability and robustness of treatment effectiveness; and informing us when research finds are especially sensitive to investigators' research design. The pitfalls include aggregating data from studies on different populations; aggregating when there is more than one underlying measure of central tendency; and emphasizing an average outcome when partitioning variance gives far more useful information.

Bibliographies as Topic↗