PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Regression Analysis”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

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

At least 145 records · Page 8Linked to original sources
PubMed1993

Regional cerebral blood flow estimation by neural network-based parametric regression analysis.

An artificial neural network (ANN) model was proposed for real-time estimation of regional cerebral blood flow (rCBF), by given head and expired air curves obtained through 133Xe inhalation. The network was constructed according to a regression model described by a linear differential equation. Experimental results compare well with those obtained by conventional curve fitting strategies, but the parameter estimation process is much simplified. A systematic procedure in developing ANN for parametric regression analysis was introduced; networks are constructed according to the selected regression model so that the obtained weights of a trained network directly represent parameters of the regression model which best fits the observed data set. Such a design-oriented methodology extends the classification-based applications of ANN to parametric regression analysis, and therefore may have more generalized applications besides rCBF estimation.

Algorithms↗
PubMed1976

Error structure of enzyme kinetic experiments. Implications for weighting in regression analysis of experimental data.

Knowledge of the error structure of a given set of experimental data is a necessary prerequisite for incisive analysis and for discrimination between alternative mathematical models of the data set. A reaction system consisting of glutathione S-transferase A (glutathione S-aryltransferase), glutathione, and 3,4-dichloro-1-nitrobenzene was investigated under steady-state conditions. It was found that the experimental error increased with initial velocity, v, and that the variance (estimated by replicates) could be described by a polynomial in v Var (v) = K0 + K1 - v + K2 - v2 or by a power function Var (v) = K0 + K1 - vK2. These equations were good approximations irrespective of whether different v values were generated by changing substrate or enzyme concentrations. The selection of these models was based mainly on experiments involving varying enzyme concentration, which, unlike v, is not considered a stochastic variable. Different models of the variance, expressed as functions of enzyme concentration, were examined by regression analysis, and the models could then be transformed to functions in which velocity is substituted for enzyme concentration owing to the proportionality between these variables. Thus, neither the absolute nor the relative error was independent of velocity, a result previously obtained for glutathione reductase in this laboratory [BioSystems 7, 101-119 (1975)]. If the experimental errors or velocities were standardized by division with their corresponding mean velocity value they showed a normal (Gaussian) distribution provided that the coefficient of variation was approximately constant for the data considered. Furthermore, it was established that the errors in the independent variables (enzyme and substrate concentrations) were small in comparison with the error in the velocity determinations. For weighting in regression analysis the inverted value of the local variance in each experimental point should be used. It was found that the assumption of proportionality between variance and valpha (where alpha is an empirically determined exponent) was a good approximation for the weighting. The value of alpha was 1.6 in the present case. The weight function was tested in the fitting of a rate equation to a kinetic-data set involving variable substrate concentrations. Recommendations are given regarding the establishment of the error structure in a general case and its application in regression analysis.

Analysis of Variance↗
PubMed2002

Segmented regression analysis of interrupted time series studies in medication use research.

Interrupted time series design is the strongest, quasi-experimental approach for evaluating longitudinal effects of interventions. Segmented regression analysis is a powerful statistical method for estimating intervention effects in interrupted time series studies. In this paper, we show how segmented regression analysis can be used to evaluate policy and educational interventions intended to improve the quality of medication use and/or contain costs.

Cost Control↗
PubMed2002

Relationship between cadmium concentration in rice and renal dysfunction in individual subjects of the Jinzu River basin determined using a logistic regression analysis.

We investigated the association between the Cd concentration in rice and renal dysfunction in individuals living in the Cd-polluted Jinzu River basin, using a logistic regression analysis. In the cases of logistic regression analysis for people (1) who had either resided in the present hamlet since birth or who had moved there from a non-polluted area and for those (2) who had resided in the present hamlet since birth, except for glucosuria in males, all partial correlation coefficients between the Cd concentration in rice and occurrence of abnormal urinary findings were statistically significant in both sexes. The allowable level of Cd concentration in rice was calculated by substituting the abnormality rates of urinary findings of the controls in the 40-49, 50-59 and 60-69 year age groups into the logistic regression formula for people (3). The value for subjects aged 50 years was 0.13 and 0.17 ppm for males and females, respectively, with regard to proteinuria and 0.15 and 0.10 ppm for males and females, respectively, with regard to proteinuria+glucosuria.

Adult↗
PubMed2007

Functional regression analysis using an F test for longitudinal data with large numbers of repeated measures.

Longitudinal data sets from certain fields of biomedical research often consist of several variables repeatedly measured on each subject yielding a large number of observations. This characteristic complicates the use of traditional longitudinal modelling strategies, which were primarily developed for studies with a relatively small number of repeated measures per subject. An innovative way to model such 'wide' data is to apply functional regression analysis, an emerging statistical approach in which observations of the same subject are viewed as a sample from a functional space. Shen and Faraway introduced an F test for linear models with functional responses. This paper illustrates how to apply this F test and functional regression analysis to the setting of longitudinal data. A smoking cessation study for methadone-maintained tobacco smokers is analysed for demonstration. In estimating the treatment effects, the functional regression analysis provides meaningful clinical interpretations, and the functional F test provides consistent results supported by a mixed-effects linear regression model. A simulation study is also conducted under the condition of the smoking data to investigate the statistical power for the F test, Wilks' likelihood ratio test, and the linear mixed-effects model using AIC.

Behavior Therapy↗
PubMed2002

Education, income inequality, and mortality: a multiple regression analysis.

OBJECTIVE: To test whether the relation between income inequality and mortality found in US states is because of different levels of formal education. DESIGN: Cross sectional, multiple regression analysis. SETTING: All US states and the District of Columbia (n=51). DATA SOURCES: US census statistics and vital statistics for the years 1989 and 1990. MAIN OUTCOME MEASURE: Multiple regression analysis with age adjusted mortality from all causes as the dependent variable and 3 independent variables-the Gini coefficient, per capita income, and percentage of people aged >/=18 years without a high school diploma. RESULTS: The income inequality effect disappeared when percentage of people without a high school diploma was added to the regression models. The fit of the regression significantly improved when education was added to the model. CONCLUSIONS: Lack of high school education accounts for the income inequality effect and is a powerful predictor of mortality variation among US states.

Adult↗
PubMed2004

Multiple linear regression analysis of the seasonal changes in the serum concentration of beta-cryptoxanthin.

Beta-cryptoxanthin (beta-CRX) is a carotenoid pigment found in Satsuma mandarin (Citrus unshiu Marc.) fruit, which is heavily produced in Japan. In this study, we evaluated the seasonal changes in the serum beta-CRX level and investigated predictors of serum beta-CRX level by multiple linear regression analysis. Blood tests and self-administered questionnaires were used every other month for one year. The subjects were healthy volunteers, 15 males and 12 females. The serum beta-CRX levels increased dramatically as the intake of Satsuma mandarin increased; the maximum increase was noted in January. Multiple linear regression analysis showed that, in males, the serum beta-CRX level could be predicted by Satsuma mandarin intake, age and the month of blood sampling; however, it was inversely associated with alcohol and smoking habits. Conversely, in females, the serum beta-CRX concentration could be predicted by Satsuma mandarin intake, the month of blood sampling and age; however, it was inversely associated with body mass index. The results of multiple linear regression analysis suggest that the serum beta-CRX levels can be used to evaluate the intake volume of Satsuma mandarin. Furthermore, beta-CRX is a useful biomarker to estimate the beneficial effects of Satsuma mandarin intake in epidemiological studies.

Adult↗
PubMed2004

Ratio variables in regression analysis can give rise to spurious results: illustration from two studies in periodontology.

OBJECTIVES: For over a century, statisticians have highlighted concerns about the inappropriate use of ratio variables in correlation and regression analysis. However, little attention has been paid to these concerns in medical and dental research. The use of ratio variables in correlation and regression analysis can give rise to spurious results due to inappropriate model specification and mathematical coupling, leading to serious misinterpretation of data and consequently to incorrect study conclusions. METHODS: Data were reanalysed from two recently published articles: one on the efficacy of guided tissue regeneration on root coverage; the other a randomised controlled trial comparing three surgical approaches in the treatment of periodontal infrabony defects. The reanalysis was performed to examine whether the assumptions behind the correlation/regression analyses have been seriously violated in these two studies, and to see if the interpretation of results is tenable. RESULTS: Use of ratio variables seriously violated the assumptions underpinning the statistical methods utilised in these two studies, and consequently the conclusions were substantially misleading. Recommendations made in these studies were not tenable. CONCLUSIONS: The reanalyses illustrate how the inappropriate use of ratio variables remains prevalent in dental research, leading to incorrect interpretation of the evidence. This emphasises the need for collaboration between clinicians and statisticians to avoid the risk of yielding erroneous conclusions from flawed statistical analyses.

Alveolar Bone Loss↗
PubMed1996

Understanding logistic regression analysis through example.

Logistic regression is a valuable statistical tool that is often used in primary care research. When researchers explore the association between a possible risk factor and a disease, they attempt to control the effects of extraneous factors (confounders) that can obscure the true association. Using logistic regression, researchers can simultaneously control for the effects of multiple confounders. When investigators use logistic regression, they make subjective decisions about which factors to include in the analysis and in the final predictive model. Critical readers must understand basic concepts of logistic regression and potential problems with its use before they can accurately interpret study results. This article uses a familiar example to explain the principles of logistic regression to make it understandable to nonstatisticians.

Coffee↗
PubMed2001

[Multivariate logistic regression analysis on the risk factors of type 2 diabetes mellitus].

OBJECTIVE: To explore the relationship between multivariate factors and type 2 diabetes mellitus. METHODS: Polymorphisms of microsatellite markers in uncoupling protein 3 (UCP3) gene, hormone-sensitive lipase (HSL) gene and protein tyrosine phosphatase-1B (PTP-1B) gene were tested in 106 patients with type 2 diabetes and 102 control subjects by performing polymerase chain reaction (PCR), polyacrylamide gel electrophoresis and silver staining. Multivariate logistic regression analysis was performed by all factors. RESULTS: Through univariate analysis, type 2 diabetes had significantly positive associations with age, systolic blood pressure (SBP), fasting insulin (FINS) level, cholesterol, triglyceride, low-density lipoprotein, apolipoprotein B (ApoB), alpha lipoprotein, UCP3 gene allele 6, UCP3 gene allele 7, HSL gene allele 9, and negative associations with UCP3 gene allele 1 and 3, HSL gene allele 5, high-density lipoprotein. The results of multivariate logistic analysis showed that UCP3 gene allele 6, UCP3 gene allele 7, SBP, ApoB, alpha lipoprotein were still positively related to type 2 diabetes, while HSL gene allele 5, high-density lipoprotein were still negatively related to type 2 diabetes. CONCLUSION: Our data showed that UCP3 gene allele 6, UCP3 gene allele 7, SBP, abnormity of plasma ApoB and alpha lipoprotein might play a role in the development of type 2 diabetes. HSL gene allele 5, high-density lipoprotein might play some protective role in the development of type 2 diabetes.

Blood Pressure↗
PubMedDate not supplied

Regression analysis of pesticide use and breast cancer incidence in California Latinas.

An evaluation of pesticide use data and breast cancer incidence rates in California Hispanic females was conducted via a regression analysis. The analysis used 1988-2000 data from the California Cancer Registry, the population-based cancer registry that monitors cancer incidence and mortality in California. It also used pesticide use data from 1970-1988 from the California Department of Pesticide Regulation. California is the leading agricultural state in the United States, and more than a quarter of all pesticides in the United States are applied there. Hispanic (Latina) females are commonly employed in agricultural operations. The authors performed regression analysis of county-level specific pesticide use data (pounds of active ingredients applied) for two classes of pesticides, organochlorines and triazine herbicides, against the breast cancer incidence rates among Latinas, controlling for age, socioeconomic status, and fertility rates, using negative binomial regression models. A total of 23,513 Latinas were diagnosed with breast cancer in California during the years 1988-1999. Risk of breast cancer was positively and significantly associated with age and socioeconomic status, and inversely and significantly associated with fertility levels. With respect to pesticides, breast cancer was positively associated with pounds of the organochlorines methoxychlor (adjusted incidence rate ratio [IRR] for highest quartile = 1.18; confidence interval [CI] = 1.03-1.35) and toxaphene (IRR = 1.16; CI = 1.01-1.34). No significant associations were found for the triazine herbicides atrazine and simazine.

Adult↗
PubMed2025

Composition-on-composition regression analysis for multi-omics integration of metagenomic data.

MOTIVATION: Compositional data are frequently encountered in many disciplines, such as in next-generation sequencing experiments widely used in biomedical studies. Regression analysis with compositional data as either responses or predictors has been well studied. However, when both responses and predictors are compositional, the inventory of analysis tools is surprisingly limited, especially in the high-dimensional setting. Among the few existing methods, most of them rely on a log-ratio transformation to move compositional data from the simplex to real numbers. Yet, a serious weakness of these methods is their failure to handle the substantial fraction of zeroes observed in data collected from next-generation sequencing experiments. RESULTS: To investigate associations between two high-dimensional multi-omics compositions, we propose a composition-on-composition (COC) regression analysis method which does not require log-ratio transformations and hence can handle zeroes in the data. To account for high dimensionality, we estimate regression coefficients using a penalized estimation equation approach. Finally, inference procedures for COC regression are also proposed. Superior performance of COC is demonstrated through both comprehensive numerical simulations and case studies. AVAILABILITY AND IMPLEMENTATION: Source R codes to implement COC method is available at https://github.com/nrios4/COC.

Regression Analysis↗
PubMed2003

Array rank order regression analysis for the detection of gene copy-number changes in human cancer.

cDNA microarray technology has been applied to the detection of DNA copy-number changes in malignant tumors. Test and control genomic DNA samples are differentially labeled and cohybridized to a spotted cDNA microarray. The ratio of test to control fluorescence intensities for each spot reflects relative gene copy number. The low signal-to-noise ratios of this assay and the variable levels of gene amplification and deletion among tumors hamper the detection of deviations from the diploid complement. We describe a regression-based statistical method to test for altered copy number on each gene and apply the technique to copy-number profiles in 10 thyroid tumors. We show that a novel transformation of fluorescence ratios into array rank order efficiently normalizes the heterogeneity among copy-number profiles and improves the reproducibility of the results. Array rank order regression analysis enhances the detection of consistent changes in gene copy number in solid tumors by cDNA microarray-based comparative genome hybridization.

Fluorescence↗
PubMed1993

Dose-response relationship between total cadmium intake and beta 2-microglobulinuria using logistic regression analysis.

The dose-response relationship between total cadmium intake and beta 2-microglobulinuria was investigated using logistic regression analysis in order to consider the effect of age on this association. The target population consisted of 1850 inhabitants of the cadmium-polluted Kakehashi River basin in Ishikawa prefecture, Japan. They were divided into 58 subgroups (27 in the men and 31 in the women) by four factors of sex, age, rice cadmium concentration and length of residence in cadmium-polluted areas. Logistic regression analysis was performed for this dose-response relationship, and both age and total cadmium intake were significantly associated with beta 2-microglobulinuria. It was confirmed that total cadmium intake had a significant association with beta 2-microglobulinuria, independent of the aging effect.

Age Factors↗
PubMed1997

[Sudden deafness: stepwise regression analysis on the correlation factors with prognosis].

To further investigate the correlation factors of the prognosis and therapeutic efficiency of sudden deafness, 83 cases (91 ears) from 1993. 1 to 1996. 4 were studied with stepwise regression analysis. The results showed that the prognosis is correlate with the age, the hearing threshold levels of both speech and high frequencies, and the types of the audiometric curve, but not with the illness duration, sex, vertigo, ipsi or bilateral, recurrence, virus infection, cardiac or cerebral vascular diseases, and the medicine or the duration of the treatment. By comparing the results obtained from different statistical analysis, it was suggested that the unification of the criteria of the diagnosis and the therapeutic efficiency as well as the application of the multiple factors regression analysis are essential for the research on sudden deafness.

Adult↗
PubMedDate not supplied

Risk group definition in children with acute myeloid leukemia by calculating individual risk factors on the basis of a multivariate stepwise Cox regression analysis.

To define risk groups in children with acute myeloid leukaemia (AML), we conducted a multivariate stepwise Cox regression analysis of three consecutive multicentre studies in East Germany. The total number of patients was 240, but cytogenetics and remission status on day 15 were routinely investigated only in the most recent study, AML-III/93 (78 patients). We derived an equation to calculate individual risk factors, determined those risk factors for all patients of study AML-III/93 and divided them into three groups with 26 patients in each. The variables in the equation were: WBC, FAB-type, auer rods, cytogenetics and response status on day 15. The event-free survival was 80% in the low risk, 55% in the intermediate risk and 15% in the high risk group. Our results strongly suggest that calculating individual risk factors on the basis of a multivariate stepwise Cox regression analysis is a useful tool in defining risk groups.

Child↗
PubMed1994

[Study on intractable factors in urinary tract infections--multiple regression analysis].

Intractable complicated urinary tract infections (UTI) are caused by host and/or bacterial factors which predispose to persistent infections and recurrent infections. It is still unknown what kind of factors are responsible for the intractable complicated UTI. With the increase of compromised host or cases with complicated UTI, the factors involved in intractable and recurrent UTI are diversified. Accurate diagnosis of the factors affecting the therapeutic effect will have more importance. The factors affecting the therapeutic effect was subjected to multiple regression analysis from both aspects of underlying disease in the urinary tract (complicated factors) and systemic conditions (compromised factors) in one hundred and ninety patients of complicated UTI admitted to our clinic. Sex, presence or absence of hydronephrosis and indwelling catheter and volume of residual urine as complicated factors and age, serum creatine value, peripheral neutrophil count, peripheral lymphocyte count, diabetic or not, whether the patient underwent major operation within 1 week or not, and serum albumin value, an indicator of malnutrition, as compromised factors were analyzed by multiple regression analysis. The presence of indwelling catheter, residual urine more than 50 ml and hypoalbuminemia less than 3 g/dl were the most determinant of the clinical efficacy in cases with complicated UTI. Interestingly, presence of residual urine more than 50 ml is considered an equally intractable factor with the presence of indwelling catheter. These factors proved to be important also as recurrent factors. In the host with these factors, UTI are often caused by resistant bacteria and indicated to be intractable also bacteriologically.

Adolescent↗