Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Multivariate statistics”

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 1,171 records · Page 65Linked to original sources
PubMed1991

Logistic-regression model for assessing portal hypertension by measuring hyaluronic acid (hyaluronan) and laminin in serum.

We earlier observed a positive correlation between portal venous pressure (PVP) and the concentration of laminin in serum. Here we investigated whether the diagnostic efficacy could be improved by considering additional analytes and application of multivariate statistical analysis. In 102 patients with fibrotic liver disease of various etiologies we measured PVP as the gradient of the wedged and free hepatic venous pressures and determined the concentrations of hyaluronic acid and laminin in serum. Regression coefficients established by logistic regression in this group were subsequently used to predict portal hypertension (PVP greater than mmHg) in an independent group of 45 patients. By comparison with the known PVP, we obtained a sensitivity of 0.83 (confidence interval: 0.63-0.93) and a specificity of 0.82 (0.61-0.93) for diagnosis of portal hypertension by means of the concentrations of hyaluronic acid and laminin in serum. Application of the model is suggested as a tool for pre-screening and monitoring patients to be subjected to assessment of portal hypertension.

Adult↗
PubMed2001

Identification of Enterococcus, Streptococcus, and Staphylococcus by multivariate analysis of proton magnetic resonance spectroscopic data from plate cultures.

A new fingerprinting technique with the potential for rapid identification of bacteria was developed by combining proton magnetic resonance spectroscopy ((1)H MRS) with multivariate statistical analysis. This resulted in an objective identification strategy for common clinical isolates belonging to the bacterial species Staphylococcus aureus, Staphylococcus epidermidis, Enterococcus faecalis, Streptococcus pneumoniae, Streptococcus pyogenes, Streptococcus agalactiae, and the Streptococcus milleri group. Duplicate cultures of 104 different isolates were examined one or more times using (1)H MRS. A total of 312 cultures were examined. An optimized classifier was developed using a bootstrapping process and a seven-group linear discriminant analysis to provide objective classification of the spectra. Identification of isolates was based on consistent high-probability classification of spectra from duplicate cultures and achieved 92% agreement with conventional methods of identification. Fewer than 1% of isolates were identified incorrectly. Identification of the remaining 7% of isolates was defined as indeterminate.

Culture Media↗
PubMedDate not supplied

The use of multivariate methods in the identification of subtypes of Alzheimer's disease: a comparison of principal components and cluster analysis.

Two contrasting multivariate statistical methods, viz., principal components analysis (PCA) and cluster analysis were applied to the study of neuropathological variations between cases of Alzheimer's disease (AD). To compare the two methods, 78 cases of AD were analyzed, each characterised by measurements of 47 neuropathological variables. Both methods of analysis revealed significant variations between AD cases. These variations were related primarily to differences in the distribution and abundance of senile plaques (SP) and neurofibrillary tangles (NFT) in the brain. Cluster analysis classified the majority of AD cases into five groups which could represent subtypes of AD. However, PCA suggested that variation between cases was more continuous with no distinct subtypes. Hence, PCA may be a more appropriate method than cluster analysis in the study of neuropathological variations between AD cases.

Aged↗
PubMed1995

Nutritional change and economic crisis in an urban Congolese community.

BACKGROUND: In 1986, the government of Congo undertook a structural adjustment programme to cope with the economic crisis. We present the results of a study whose objectives were to assess the evolution of nutritional status of an urban community between 1986 and 1991 and to identify specific groups for which the nutritional status may have worsened. METHODS: Two cross-sectional surveys were carried out on representative samples of Brazzaville children < 6 years old: 2295 children were surveyed in 1986 and 2373 in 1991. Anthropometric assessment of nutritional status was performed. For children, weight-for-height and height-for-age indices were used according to WHO recommendations. Wasting and stunting were respectively defined as indices under -2 z-scores. Body mass index of mothers was calculated and risk of chronic energy deficiency (CED) was defined as < 18.5 kg/m2. Socioeconomic data relative to the households were also collected. Multivariate statistical methods were used to obtain adjusted estimates of nutritional changes in the community. RESULTS: Data analysis led to several converging results: increase in the percentage of low birthweight (10.2% in 1985 versus 18.7% in 1990), increase in the percentage of CED (from 7.9% to 10.5%), and increase in the prevalence of wasting (from 2.9% to 4.2%). By contrast, the overall prevalence of stunting decreased from 13.9% to 11.0%. After statistical adjustment, the factors found to influence the evolution of anthropometric status were: age of child, age of mother, schooling of mother and household characteristics such as number of preschool children, economic level and head of household's occupation. CONCLUSIONS: The study enables the negative effects of the economic crisis to be quantified. Body mass index is shown to be sensitive to economic changes. It could be recommended as a possible indicator for monitoring the nutritional status at population level. The results also call for a new impetus in preventive health programmes and the implementation of nutritional surveillance activities.

Adolescent↗
PubMed2004

Noninvasive measurement of glucose by metabolic heat conformation method.

BACKGROUND: We developed a method, called the metabolic heat conformation (MHC) method, for the noninvasive measurement of blood glucose. The MHC method involves the measurement of physiologic indices related to metabolic heat generation and local oxygen supply, which correspond to the glucose concentration in the local blood supply. METHODS: We used noninvasive thermal and optical sensors on the fingertip of an individual to measure thermal generation, blood flow rate, hemoglobin (Hb) concentration, and oxyhemoglobin concentration. The calibration model incorporates mathematical procedures to convert signals from the sensor pickup to final glucose concentrations. The mathematical procedures are multivariate statistical analyses, involving values from sensor signals, polynomials from various values, regression analyses of individual patients, and cluster analyses of patient groups. The glucose value is calculated for each patient measurement, applying one of the clusters by discriminant analysis. RESULTS: Regression analysis was performed to compare the noninvasive method with the hexokinase method, using 127 data points (109 data points from diabetic patients, 18 data points from nondiabetic patients) with glucose concentrations ranging from 3.0 to 22.5 mmol/L (54-405 mg/dL). The correlation coefficient (r) was 0.91. Reproducibility was measured for healthy fasting persons; the CV was 6% at 5.56 mmol/L (100 mg/dL). CONCLUSIONS: These data provide preliminary evidence that the MHC method can be used to estimate blood glucose concentrations noninvasively.

Blood Glucose↗
PubMed2005

Predictors of improvement in hemoglobin concentration among toddlers enrolled in the Massachusetts WIC Program.

OBJECTIVE: Nutrition supplementation programs are generally targeted to those members of the population who are thought to be at risk of an undesirable outcome, but not all who participate in such programs respond to them. We sought to identify determinants of improvement in hemoglobin concentration among young children in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC). DESIGN: We conducted an observational study using data from 9,930 children who were enrolled in the Massachusetts WIC program and had data available on their hemoglobin values at both 1 and 2 years of age. Predictors of change in hemoglobin concentration between these ages were studied using multivariate statistical modeling. RESULTS: Overall, hemoglobin concentrations increased from age 1 to 2 years in those who had been breastfed 25 or more weeks (P < .0001) and were female (P < .01), and decreased with increasing weight at 1 year of age (P < .001). The determinants of change in hemoglobin concentration differed from the determinants of hemoglobin concentration at age 1 year. CONCLUSIONS: The analytical approach used here could be extended to identify subgroups of WIC participants likely to improve in other outcomes. If current efforts to increase the duration of breastfeeding among WIC participants are successful, the importance of WIC in improving hemoglobin concentration among young children also will increase.

Anemia, Iron-Deficiency↗
PubMed2005

The associations between leisure-time physical activity and inflammatory and coagulation markers related to cardiovascular disease: the ATTICA Study.

BACKGROUND: As coronary heart disease is increasingly seen as an inflammation process, we evaluated the hypothesis whether physical activity reduces coronary heart disease risk by modifying the levels of inflammatory and coagulation markers. METHODS: From May 2001 to December 2002, we randomly enrolled 1524 adult men and 1518 women, without any evidence of cardiovascular disease, stratified by age-gender (census 2001), from the greater area of Athens, Greece. We assessed the relationship between self-reported physical activity status and inflammation markers (high sensitivity C-reactive protein, serum amyloid-A, fibrinogen, interleukin-6, tumor necrosis factor-alpha, and white blood cell counts), after taking into account the effect of several confounders. RESULTS: Eight hundred seventy-four (57%) of men and 903 (59%) of women were classified as sedentary. Multivariate statistical analysis after adjustment for gender, age, smoking habits, body mass index, total cholesterol, blood glucose, and systolic and diastolic blood pressure levels showed that participants devoted to high physical activity (>7 kcal/min expended) had 29% lower levels of C-reactive protein, 19% of white blood cell counts, 22% lower concentrations of amyloid-A, 20% lower levels of tumor necrosis factor-alpha, 32% of interleukin-6, and 11% of fibrinogen (all P<0.05) as compared to those who were devoted to sedentary life. CONCLUSIONS: Our findings suggest that the adoption of a physically active lifestyle modifies the inflammation process in healthy individuals.

Adult↗
PubMed1999

Comparison of API 20NE and Biolog GN identification systems assessed by techniques of multivariate analyses.

The increasing use of commercial multitest systems for identification of environmental bacteria creates the problem of how to compare the identification results obtained from different systems. The limited use of species designations in such comparisons is caused by low usage of environmental bacteria in the development of commercial identification schemes. Two multivariate statistical methods, the Mantel's test and the co-inertia analysis, were applied to analyze data derived from the Biolog GN and the API 20NE systems of identification for 50 environmental bacterial strains. We found these two methods to be useful for revealing the relationship between the two sets of numerical taxonomic traits. Both of these methods showed that the distances according to the Biolog GN results between the studied strains were related to those derived from the API 20NE results, despite the differences in the test sets of the two systems. In addition, the co-inertia analysis allowed us to visualise the relationships between classifications of strains derived from the two identification systems and, simultaneously, to estimate the contribution of particular tests to the differentiation of bacterial strains.

Bacterial Typing Techniques↗
PubMed1994

Cognitive impairment in human chronic Chagas' disease.

UNLABELLED: We proposed to investigate subclinical cognitive impairment secondary to chronic Chagas' disease (CCD). No similar study was previously done. The neuropsychological performance of 45 chronic Chagasic patients and 26 matched controls (age, education place and years of residency in endemic area) was compared using the Mini Mental State Exam (MMSE), Weschler Memory Scale (WMS) and the Weschler Adult Intelligent Scale (WAIS). Non-parametric tests and Chi2 were used to compare group means and multivariate statistics in two way frequency tables for measures of independence and association of categorical variables with the disease. RESULTS: Chagasic patients showed lower MMSE scores (p < .004), poor orientation (p < .004), and attention (p < .007). Lower WMS MQ were associated with CCD (Chi2 5.9; p < .01; Fisher test p < .02). Lower WAIS IQ were associated with CCD (Chi2 6.3, p < .01; Fisher test p < .01) being the digit symbol (p < .03), picture completion (p < .03), picture arrangement (p < .01) and object assembly (p < .03) subtests the most affected. The impairment in non-verbal reasoning, speed of information processing, problem solving, learning and sequencing observed in chronic Chagas disease patients resembles the cognitive dysfunction associated with white matter disease.

Adult↗
PubMed1998

Bacterial corneal ulcer: a multivariate study.

A retrospective clinicomicrobiological review of 314 patients with bacterial corneal ulcers from January 1982 to December 1992 was performed. Multivariate statistical analysis was done with multiple logistic regression using PROC LOGIST of SAS statistical software. Positive cultures were grown from 134 (42.7%) of the patients. Pseudomonas aeruginosa, staphylococci, and Acinetobacter spp. were the most frequent pathogens. Significant associations between contact lens use and P. aeruginosa (odds ratio, OR 8.16), between previous herpes simplex keratitis and Streptococcus spp. (OR 18.2) were found. Acinetobacter spp. occurred more frequently in eyes with burn and/or lagophthalmos (OR 13.1/26.2). Staphylococcus aureus was associated with trauma (OR 6.27) and age under 50 (OR 5.08-13.6). Nonpseudomonal gram-negative bacilli were associated with age over 50 (OR 3.24). Drug sensitivity tests for these isolated microorganisms showed that vancomycin and ceftazidime were the most effective agents.

Adolescent↗
PubMed2005

Environmental prognostics: an integrated model supporting lysosomal stress responses as predictive biomarkers of animal health status.

The potential prognostic use of lysosomal reactions to environmental pollutants is explored in relation to predicting animal health in marine mussels, based on diagnostic biomarker data. Cellular lysosomes are already known to accumulate many metals and organic xenobiotics and the lysosomal accumulation of the carcinogenic polycyclic aromatic hydrocarbon 3-methylcholanthrene (3-MC) is demonstrated here in the hepatopancreatic digestive cells and ovarian oocytes of the blue mussel. Lysosomal membrane integrity or stability appears to be a generic indicator of cellular well-being in eukaryotes; and in bivalve molluscs it is correlated with total oxygen and nitrogen radical scavenging capacity (TOSC), protein synthesis, scope for growth and larval viability; and inversely correlated with DNA damage (micronuclei), as well as lysosomal swelling (volume density), lipidosis and lipofuscinosis, which are all characteristic of failed or incomplete autophagy. Integration of multiple biomarker data is achieved using multivariate statistics and then mapped onto "health status space" by using lysosomal membrane stability as a measure of cellular well-being. This is viewed as a crucial step towards the derivation of explanatory frameworks for prediction of pollutant impact on animal health; and has facilitated the development of a conceptual mechanistic model linking lysosomal damage and autophagic dysfunction with injury to cells, tissues and the whole animal. This model has also complemented the creation and use of a cell-based bioenergetic computational model of molluscan hepatopancreatic cells that simulates lysosomal and cellular reactions to pollutants. More speculatively, the use of coupled empirical measurements of biomarker reactions and modelling is proposed as a practical approach to the development of an operational toolbox for predicting the health of the environment.

Animal Structures↗
PubMed1991

Bleeding complications to oral anticoagulant therapy: multivariate analysis of 1010 treatment years in 551 outpatients.

One thousand and ten patient years of oral anticoagulant therapy with vitamin-K-antagonists were reviewed with regard to major bleeding complications. The incidence of bleeding that necessitated hospital admission was 2.7% per year (95% confidence limits, 1.7-3.7%). The major source of bleeding was the alimentary tract, whereas no cases of intracranial bleeding were found. Various factors with potential effects on the bleeding risk were evaluated by multivariate statistical analysis, and the following independent risk factors were identified: age greater than 75 years and hypertension increased the bleeding risk by 10.5% and 4.5%, respectively. Each recorded prothrombin value significantly below the therapeutic range increased the bleeding risk by 3.9%, and each year of treatment increased the risk by 2.0%. These figures may be used to estimate the risk of major bleeding in an individual patient. Current treatment with thiazide diuretics was found to increase the bleeding risk by 5.2%. However, this observation requires further documentation and analysis. Although no lethal episodes of bleeding occurred, the developing field of indications for oral anticoagulant therapy should be considered on the basis of a continuous substantial risk of major bleeding.

Adolescent↗
PubMed1999

Increased lipoprotein(a) is an important risk factor for venous thromboembolism in childhood.

BACKGROUND: Serum levels of lipoprotein(a) [Lp(a)] are determined largely by genetic variation in the gene encoding for apolipoprotein(a) [apo(a)], the specific protein component of Lp(a) that is very homologous to plasminogen. High plasma levels of Lp(a) increase the risk for premature atherosclerotic vessel diseases. We investigated the little-characterized role of Lp(a) as a risk factor for venous thromboembolic diseases, alone and in conjunction with established thrombophilic risk factors of proteins regulating blood coagulation and fibrinolysis. METHODS AND RESULTS: Serum levels of Lp(a) and lipids, protein C, protein S, and antithrombin, as well as the size of apo(a) isoforms and the presence of the factor V:Q(506) mutation, were determined in 186 consecutively admitted children from neonates to 18 years old with a history of venous thrombosis and in 186 age- and disease-matched control subjects. Children with a history of venous thrombosis had a significantly higher median Lp(a) level (19 versus 4.4 mg/dL) than control subjects. The risk for thromboembolic events in children with Lp(a) levels in the upper quartile, ie, >30 mg/dL, was 7.2 (95% CI, 3.7 to 14.5). The size of apo(a) isoforms was inversely related to Lp(a) levels and to the risk for thromboembolic events. Compared with the highest quartile of the apo(a) size distribution, the lowest quartile was associated with a risk of 8.2. In addition, multivariate statistical analysis gives evidence that the factor V:Q(506) mutation (OR/CI, 2.8/1.6 to 4.9), protein C (OR/CI, 6.5/2.1 to 19), and antithrombin deficiency (OR/CI, 10.4/1.2 to 90) were independent risk factors of childhood venous thrombosis. Coincidence of elevated Lp(a) with factor V:Q(506) mutation or deficiencies of protein C or antithrombin further increased the risk for thromboembolic events to 8.4. CONCLUSIONS: Lp(a) >30 mg/dL is a risk factor for venous thromboembolism in childhood. Lp(a) measurements should be included in the screening of causal factors in children with venous thromboembolic events.

Activated Protein C Resistance↗
PubMed2001

Assessing patients with possible heart disease using scores.

Multivariable analysis of clinical and exercise test data has the potential to become a useful tool for assisting in the diagnosis of coronary artery disease, assessing prognosis, and reducing the cost of evaluating patients with suspected coronary disease. Since general practitioners are functioning as gatekeepers and decide which patients must be referred to the cardiologist, they need to use the basic tools they have available (i.e. history, physical examination and the exercise test), in an optimal fashion. Scores derived from multivariable statistical techniques considering clinical and exercise data have demonstrated superior discriminating power compared with simple classification of the ST response. In addition, by stratifying patients as to probability of disease and prognosis, they provide a management strategy. While computers, as part of information management systems, can run complicated equations and derive these scores, physicians are reluctant to trust them. Thus, these scores have been represented as nomograms or simple additive tables so physicians are comfortable with their application. Their results have also been compared with physician judgment and found to estimate the presence of coronary disease and prognosis as well as expert cardiologists and often better than nonspecialists. However, the discriminating power of specific variables from the medical history and exercise test remains unclear because of inadequate study design and differences in study populations. Should expired gases be substituted for estimated metabolic equivalents (METs)? Should ST/heart rate (HR) index be used instead of putting these measurements separately into the models? Should right-sided chest leads and HR in recovery be considered? There is a need for further evaluation of these routinely obtained variables to improve the accuracy of prediction algorithms especially in women. The portability and reliability of these equations must be demonstrated since access to specialised care must be safe-guarded. Hopefully, sequential assessment of the clinical and exercise test data and application of the newer generation of multivariable equations can empower the clinician to assure the cardiac patient access to appropriate and cost-effective cardiological care.

Adult↗
PubMed2002

Combining location and expression data for principled discovery of genetic regulatory network models.

We develop principled methods for the automatic induction (discovery) of genetic regulatory network models from multiple data sources and data modalities. Models of regulatory networks are represented as Bayesian networks, allowing the models to compactly and robustly capture probabilistic multivariate statistical dependencies between the various cellular factors in these networks. We build on previous Bayesian network validation results by extending the validation framework to the context of model induction, leveraging heuristic simulated annealing search algorithms and posterior model averaging. Using expression data in isolation yields results inconsistent with location data so we incorporate genomic location data to guide the model induction process. We combine these two data modalities by allowing location data to influence the model prior and expression data to influence the model likelihood. We demonstrate the utility of this approach by discovering genetic regulatory models of thirty-three variables involved in S. cerevisiae pheromone response. The models we automatically generate are consistent with the current understanding regarding this regulatory network, but also suggest new directions for future experimental investigation.

Bayes Theorem↗
PubMed2006

Using multivariate analyses and GIS to identify pollutants and their spatial patterns in urban soils in Galway, Ireland.

Galway is a small but rapidly growing tourism city in western Ireland. To evaluate its environmental quality, a total of 166 surface soil samples (0-10 cm depth) were collected from parks and grasslands at the density of 1 sample per 0.25 km2 at the end of 2004. All samples were analysed using ICP-AES for the near-total concentrations of 26 chemical elements. Multivariate statistics and GIS techniques were applied to classify the elements and to identify elements influenced by human activities. Cluster analysis (CA) and principal component analysis (PCA) classified the elements into two groups: the first group predominantly derived from natural sources, the second being influenced by human activities. GIS mapping is a powerful tool in identifying the possible sources of pollutants. Relatively high concentrations of Cu, Pb and Zn were found in the city centre, old residential areas, and along major traffic routes, showing significant effects of traffic pollution. The element As is enriched in soils of the old built-up areas, which can be attributed to coal and peat combustion for home heating. Such significant spatial patterns of pollutants displayed by urban soils may imply potential health threat to residents of the contaminated areas of the city.

Air Pollution↗
PubMed1996

Rheumatoid arthritis lung disease. Determinants of radiographic and physiologic abnormalities.

OBJECTIVE: To determine the prevalence and important clinical predictors of radiographic and physiologic abnormalities indicative of rheumatoid arthritis interstitial lung disease (RA-ILD). METHODS: An unselected cohort of patients with a confirmed diagnosis of RA and known lung disease were identified (n = 336) and evaluated for RA disease activity and severity. Outcomes included abnormalities determined by the pulmonary function tests of forced vital capacity (FVC) and diffusion capacity for carbon monoxide (DLco), and/or chest radiographic findings of interstitial infiltrates. We used multivariable statistical modeling to determine the independent significance of cigarette smoking and other RA-specific factors on the pulmonary abnormalities of interest. RESULTS: At least 1 of the 3 abnormal findings was identified by pulmonary tests in 32.4% of all patients. These abnormal findings included an FVC < 80% of predicted in 42 patients, a DLco < 80% of predicted in 64 patients, and evidence of radiographic interstitial infiltrates in 40 patients. After statistical adjustment for confounding factors, pack-years of cigarette smoking remained a significant predictor of low DLco (beta = -0.07, 95% confidence interval [95% CI] -0.09, -0.04), low FVC (beta = -0.003, 95% CI -0.006, -0.0004), and interstitial abnormalities on chest radiograph (odds ratio for > or = 25 pack-years = 3.76, 95% CI 1.59, 8.88). The Health Assessment Questionnaire (HAQ) Disability Index (DI) was also an important risk factor for the decline in both the DLco (beta = -1.15, 95% CI -2.00, -0.30) and FVC (beta = -0.23, 95% CI -0.32, -0.13). CONCLUSION: Although RA disease activity/severity (particularly as defined by the HAQ DI) was important, smoking was the most consistent independent predictor of radiographic and physiologic abnormalities suggestive of ILD in RA.

Adult↗
PubMed2004

A morphometric analysis of maxillary molar crowns of Middle-Late Pleistocene hominins.

This study explores the significance of shape differences in the maxillary first molar crowns of Neandertals and anatomically modern humans. It uses morphometric analysis to quantify these differences and to investigate how the orientation of major cusps, relative cusp base areas and occlusal polygon area influence crown shape. The aims of this study were to 1) quantify these data to test whether the tooth shapes of Neandertals and anatomically modern humans differ significantly and 2) to explore if either of the shapes is derived relative to earlier fossil hominins. Data were collected from digital occlusal photographs using image-processing software. Cusp angles, relative cusp base areas and occlusal polygon areas were measured on Neandertals (n=15), contemporary modern humans (n=62), Upper Paleolithic humans (n=6), early anatomically modern humans (n=3) and Homo erectus (n=3). Univariate and multivariate statistical tests were used to evaluate the differences between contemporary modern humans and Neandertals, while the much sparser data sets from the other fossil samples were included primarily for comparison. Statistically significant differences reflecting overall crown shape and internal placement of the crown apices were found. Neandertals are distinguished from contemporary humans by possessing maxillary first molars that 1) are markedly skewed; 2) possess a narrower distal segment of the occlusal polygon compared to the mesial segment; 3) possess a significantly smaller metacone and a significantly larger hypocone; and 4) possess a significantly smaller relative occlusal polygon area reflecting internally placed cusps. Differences in relative cusp base areas of the hypocone and metacone may contribute to the shape differences observed in Neandertals. However, early anatomically modern humans possessing a pattern of relative cusp base areas similar to Neandertals lack their unusual shape. That the morphology observed in non-Neandertal fossil hominins is more anatomically modern human-like than Neandertal-like, suggests that this distinctive morphology may be derived in Neandertals.

Animals↗