Multivariate Analysis

To develop criteria for the classification of fibromyalgia, we studied 558 consecutive patients: 293 patients with fibromyalgia and 265 control patients. Interviews and examinations were performed by trained, blinded assessors. Control patients for the group with primary fibromyalgia were matched for age and sex, and limited to patients with disorders that could be confused with primary fibromyalgia. Control patients for the group with secondary-concomitant fibromyalgia were matched for age, sex, and concomitant rheumatic disorders. Widespread pain (axial plus upper and lower segment plus left- and right-sided pain) was found in 97.6% of all patients with fibromyalgia and in 69.1% of all control patients. The combination of widespread pain and mild or greater tenderness in ⩾ 11 of 18 tender point sites yielded a sensitivity of 88.4% and a specificity of 81.1%. Primary fibromyalgia patients and secondary-concomitant fibromyalgia patients did not differ statistically in any major study variable, and the criteria performed equally well in patients with and those without concomitant rheumatic conditions. The newly proposed criteria for the classification of fibromyalgia are 1) widespread pain in combination with 2) tenderness at 11 or more of the 18 specific tender point sites. No exclusions are made for the presence of concomitant radiographic or laboratory abnormalities. At the diagnostic or classification level, the distinction between primary fibromyalgia and secondary-concomitant fibromyalgia (as defined in the text) is abandoned.

Among the 295 patients, 180 had a poorprognosis signature and 115 had a good-prognosis signature, and the mean (±SE) overall 10-year survival rates were 54.6±4.4 percent and 94.5±2.6 percent, respectively. At 10 years, the probability of remaining free of distant metastases was 50.6±4.5 percent in the group with a poor-prognosis signature and 85.2±4.3 percent in the group with a good-prognosis signature. The estimated hazard ratio for distant metastases in the group with a poor-prognosis signature, as compared with the group with the good-prognosis signature, was 5.1 (95 percent confidence interval, 2.9 to 9.0; P<0.001). This ratio remained significant when the groups were analyzed according to lymph-node status. Multivariable Cox regression analysis showed that the prognosis profile was a strong independent factor in predicting disease outcome.

Background-Separate multivariable risk algorithms are commonly used to assess risk of specific atherosclerotic cardiovascular disease (CVD) events, ie, coronary heart disease, cerebrovascular disease, peripheral vascular disease, and heart failure. The present report presents a single multivariable risk function that predicts risk of developing all CVD and of its constituents. Methods and Results-We used Cox proportional-hazards regression to evaluate the risk of developing a first CVD event in 8491 Framingham study participants (mean age, 49 years; 4522 women) who attended a routine examination between 30 and 74 years of age and were free of CVD. Sex-specific multivariable risk functions ("general CVD" algorithms) were derived that incorporated age, total and high-density lipoprotein cholesterol, systolic blood pressure, treatment for hypertension, smoking, and diabetes status. We assessed the performance of the general CVD algorithms for predicting individual CVD events (coronary heart disease, stroke, peripheral artery disease, or heart failure). Over 12 years of follow-up, 1174 participants (456 women) developed a first CVD event. All traditional risk factors evaluated predicted CVD risk (multivariable-adjusted PϽ0.0001). The general CVD algorithm demonstrated good discrimination (C statistic, 0.763 [men] and 0.793 [women]) and calibration. Simple adjustments to the general CVD risk algorithms allowed estimation of the risks of each CVD component. Two simple risk scores are presented, 1 based on all traditional risk factors and the other based on non-laboratory-based predictors. Conclusions-A sex-specific multivariable risk factor algorithm can be conveniently used to assess general CVD risk and risk of individual CVD events (coronary, cerebrovascular, and peripheral arterial disease and heart failure). The estimated absolute CVD event rates can be used to quantify risk and to guide preventive care. (Circulation. 2008;117: 743-753.)

Using data from India, we estimate the relationship between household wealth and children &#x27;s school enrollment. We proxy wealth by constructing a linear index from asset ownership indicators, us-ing principal-components analysis to derive weights. In Indian data this ...

In this article, we present FactoMineR an R package dedicated to multivariate data analysis. The main features of this package is the possibility to take into account different types of variables (quantitative or categorical), different types of structure on the data (a partition on the variables, a hierarchy on the variables, a partition on the individuals) and finally supplementary information (supplementary individuals and variables). Moreover, the dimensions issued from the different exploratory data analyses can be automatically described by quantitative and/or categorical variables. Numerous graphics are also available with various options. Finally, a graphical user interface is implemented within the Rcmdr environment in order to propose an user friendly package.

Despite multiple disparate prognostic risk analysis systems for evaluating clinical outcome for patients with myelodysplastic syndrome (MDS), imprecision persists with such analyses. To attempt to improve on these systems, an International MDS Risk Analysis Workshop combined cytogenetic, morphological, and clinical data from seven large previously reported risk-based studies that had generated prognostic systems. A global analysis was performed on these patients, and critical prognostic variables were re-evaluated to generate a consensus prognostic system, particularly using a more refined bone marrow (BM) cytogenetic classification. Univariate analysis indicated that the major variables having an impact on disease outcome for evolution to acute myeloid leukemia were cytogenetic abnormalities, percentage of BM myeloblasts, and number of cytopenias; for survival, in addition to the above, variables also included age and gender. Cytogenetic subgroups of outcome were as follows: &quot;...

We present here both linear regressions and multivariate analyses correlating three global neuropsychological tests with a number of structural and neurochemical measurements performed on a prospective series of 15 patients with Alzheimer's disease and 9 neuropathologically normal subjects. The statistical data show only weak correlations between psychometric indices and plaques and tangles, but the density of neocortical synapses measured by a new immunocytochemical/densitometric technique reveals very powerful correlations with all three psychological assays. Multivariate analysis by stepwise regression produced a model including midfrontal and inferior parietal synapse density, plus inferior parietal plaque counts with a correlation coefficient of 0.96 for Mattis's Dementia Rating Scale. Plaque density contributed only 26% of that strength.

Diffuse large B-cell lymphoma (DLBCL) can be divided into prognostically important subgroups with germinal center Bcell-like (GCB), activated B-cell-like (ABC), and type 3 gene expression profiles using a cDNA microarray. Tissue microarray (TMA) blocks were created from 152 cases of DLBCL, 142 of which had been successfully evaluated by cDNA microarray (75 GCB, 41 ABC, and 26 type 3). Sections were stained with antibodies to CD10, bcl-6, MUM1, FOXP1, cyclin D2, and bcl-2. Expression of bcl-6 (P < .001) or CD10 (P ‫؍‬ .019) was associated with better overall survival (OS), whereas expression of MUM1 (P ‫؍‬ .009) or cyclin D2 (P < .001) was associated with worse OS. Cases were subclassified using CD10, bcl-6, and MUM1 expression, and 64 cases (42%) were considered GCB and 88 cases (58%) non-GCB. The 5-year OS for the GCB group was 76% compared with only 34% for the non-GCB group (P < .001), which is similar to that reported using the cDNA microarray. Bcl-2 and cyclin D2 were adverse predictors in the non-GCB group. In multivariate analysis, a high International Prognostic Index score (3-5) and the non-GCB phenotype were independent adverse predictors (P < .0001). In summary, immunostains can be used to determine the GCB and non-GCB subtypes of DLBCL and predict survival similar to the cDNA microarray. (Blood. 2004;103:275-282)

Nonparametric multivariate analysis of ecological data using permutation tests has two main challenges: (1) to partition the variability in the data according to a complex design or model, as is often required in ecological experiments, and (2) to base the analysis on a multivariate distance measure (such as the semimetric Bray-Curtis measure) that is reasonable for ecological data sets. Previous nonparametric methods have succeeded in one or other of these areas, but not in both. A recent contribution to Ecological Monographs by Legendre and Anderson, called distance-based redundancy analysis (db-RDA), does achieve both. It does this by calculating principal coordinates and subsequently correcting for negative eigenvalues, if they are present, by adding a constant to squared distances. We show here that such a correction is not necessary. Partitioning can be achieved directly from the distance matrix itself, with no corrections and no eigenanalysis, even if the distance measure used is semimetric. An ecological example is given to show the differences in these statistical methods. Empirical simulations, based on parameters estimated from real ecological species abundance data, showed that db-RDA done on multifactorial designs (using the correction) does not have type 1 error consistent with the significance level chosen for the analysis (i.e., does not provide an exact test), whereas the direct method described and advocated here does.

Background: The dramatic progress in sequencing technologies offers unprecedented prospects for deciphering the organization of natural populations in space and time. However, the size of the datasets generated also poses some daunting challenges. In particular, Bayesian clustering algorithms based on pre-defined population genetics models such as the STRUCTURE or BAPS software may not be able to cope with this unprecedented amount of data. Thus, there is a need for less computer-intensive approaches. Multivariate analyses seem particularly appealing as they are specifically devoted to extracting information from large datasets. Unfortunately, currently available multivariate methods still lack some essential features needed to study the genetic structure of natural populations.

for prognostication and therapeutic decision making. JAMA 2000; 284: 835-842. When a patient presents with an acute STEMI, the optimal treatment is usually clear --proceed to PCI immediately. The situation is less clear with unstable angina and NSTEMI's, a heterogeneous group of acute coronary syndromes with a wide spectrum of risk for death and cardiac ischemic events. Prognostication of risk in these patients is critical as it can dictate optimal therapy and location of treatment (outpatient vs. hospital ward vs. CCU). Using the database of the TIMI 11B and ESSENCE trials (both phase III, international, randomized, double-blind trials which compared treatment with unfractionated heparin vs. enoxaparin in patients with UA/NSTEMI), twelve variables at presentation were examined with respect to a composite end-point of all-cause mortality, new or recurrent MI, or severe recurrent ischemia prompting urgent revascularization through 14 days. Seven of the 12 variables were statistically significant and formed the final set of predictor variables: 1) age greater than 65; 2) greater than 3 risk factors for CAD (family history, HTN, hyperlipidemia, DM, current smoker); 3) prior coronary stenosis of greater than 50%; 4) ST-segment deviation on EKG at presentation; 5) at least 2 anginal events in the prior 24 hours; 6) use of aspirin in the prior 7 days; 7) elevated serum cardiac markers. Event rates increased significantly as the TIMI risk score increased (0/1: 4.7%, 2: 8.3%, 3: 13.2%, 4: 19.9%, 5: 26.2%, 6/7: 40.9%). Three validation cohorts were used to validate these findings.

Background: High-risk human papillomaviruses (HPVs) are etiologic agents for anogenital tract cancers and have been detected in head and neck squamous cell carcinomas (HNSCCs). We investigated, retrospectively, an etiologic role for HPVs in a large series of patients with HNSCC. Methods: Tumor tissues from 253 patients with newly diagnosed or recurrent HNSCC were tested for the presence of HPV genome by use of polymerase chain reaction (PCR)-based assays, Southern blot hybridization, and in situ hybridization. The viral E6 coding region was sequenced to confirm the presence of tumor-specific viral isolates. Exons 5-9 of the TP53 gene were sequenced from 166 specimens. The hazard of death from HNSCC in patients with and without HPVpositive tumors was determined by proportional hazards regression analysis. Results: HPV was detected in 62 (25%) of 253 cases (95% confidence interval [CI] = 19%-30%). Highrisk, tumorigenic type HPV16 was identified in 90% of the HPV-positive tumors. HPV16 was localized specifically by in situ hybridization within the nuclei of cancer cells in preinvasive, invasive, and lymph node disease. Southern blot hybridization patterns were consistent with viral integration. Poor tumor grade (odds ratio [OR] = 2.4; 95% CI = 1.2-4.9) and oropharyngeal site (OR = 6.2; 95% CI = 3.1-12.1) independently increased the probability of HPV presence. As compared with HPV-negative oropharyngeal cancers, HPVpositive oropharyngeal cancers were less likely to occur among moderate to heavy drinkers (OR = 0.17; 95% CI = 0.05-0.61) and smokers (OR = 0.16; 95% CI = 0.02-1.4), had a characteristic basaloid morphology (OR = 18.7; 95% CI = 2.1-167), were less likely to have TP53 mutations (OR = 0.06; 95% CI = 0.01-0.36), and had improved disease-specific survival (hazard ratio [HR] = 0.26; 95% CI = 0.07-0.98). After adjustment for the presence of lymph node disease (HR = 2.3; 95% CI = 1.4-3.8), heavy alcohol consumption (HR = 2.6; 95% CI = 1.4-4.7), and age greater than 60 years old (HR = 1.4; 95% CI = 0.8-2.3), all patients with HPV-positive tumors had a 59% reduction in risk of death from cancer when compared with HPV-negative HNSCC patients (HR = 0.41; 95% CI = 0.20-0.88). Conclusions: These data extend recent molecular and epidemiologic studies and strongly suggest that HPV-positive oropharyngeal cancers comprise a distinct molecular, clinical, and pathologic disease entity that is likely causally associated with HPV infection and that has a markedly improved prognosis. [J Natl Cancer Inst 2000; 92:709-20] Head and neck squamous cell carcinoma (HNSCC) is a disease largely attributed to environmental exposures. Tobacco use and alcohol consumption are well-established risk factors . However, a small proportion (15%-20%) of HNSCCs occurs in Tumors positive for HPV16 by PCR, for which at least 2-5 g of tumor DNA were available, were examined for HPV16 sequences by Southern blot hybridization of unamplified tumor DNAs. Full-length HPV16 genomic DNA (7905 bp) in pGEM II (Promega Corp., Madison, WI) was gel purified and labeled to a high specific activity with [␣-32 P]deoxycytidine triphosphate by the random primer method (Random Primers DNA Labeling System; Life Technologies, Inc.

To determine whether graft-versus-leukemia (GVL) reactions are important in preventing leukemia recurrence after bone marrow transplantation, we studied 2,254 persons receiving HLA-identical sibling bone marrow transplants for acute myelogenous leukemia (AML) in first remission, acute lymphoblastic leukemia (ALL) in first remission, and chronic myelogenous leukemia (CML) in first chronic phase. Four groups were investigated in detail: recipients of non--T-cell depleted allografts without graft-versus-host disease (GVHD), recipients of non--T-cell depleted allografts with GVHD, recipients of T-cell depleted allografts, and recipients of genetically identical twin transplants. Decreased relapse was observed in recipients of non--T-cell depleted allografts with acute (relative risk 0.68, P = .03), chronic (relative risk 0.43, P = .01), and both acute and chronic GVDH (relative risk 0.33, P = .0001) as compared with recipients of non--T-cell depleted allografts without GVHD. These data ...

Background-The Lyon Diet Heart Study is a randomized secondary prevention trial aimed at testing whether a Mediterraneantype diet may reduce the rate of recurrence after a first myocardial infarction. An intermediate analysis showed a striking protective effect after 27 months of follow-up. This report presents results of an extended follow-up (with a mean of 46 months per patient) and deals with the relationships of dietary patterns and traditional risk factors with recurrence. Methods and Results-Three composite outcomes (COs) combining either cardiac death and nonfatal myocardial infarction (CO 1), or the preceding plus major secondary end points (unstable angina, stroke, heart failure, pulmonary or peripheral embolism) (CO 2), or the preceding plus minor events requiring hospital admission (CO 3) were studied. In the Mediterranean diet group, CO 1 was reduced (14 events versus 44 in the prudent Western-type diet group, Pϭ0.0001), as were CO 2 (27 events versus 90, Pϭ0.0001) and CO 3 (95 events versus 180, Pϭ0.0002). Adjusted risk ratios ranged from 0.28 to 0.53. Among the traditional risk factors, total cholesterol (1 mmol/L being associated with an increased risk of 18% to 28%), systolic blood pressure (1 mm Hg being associated with an increased risk of 1% to 2%), leukocyte count (adjusted risk ratios ranging from 1.64 to 2.86 with count Ͼ9ϫ10 9 /L), female sex (adjusted risk ratios, 0.27 to 0.46), and aspirin use (adjusted risk ratios, 0.59 to 0.82) were each significantly and independently associated with recurrence. Conclusions-The protective effect of the Mediterranean dietary pattern was maintained up to 4 years after the first infarction, confirming previous intermediate analyses.

Elevated serum phosphorus is a predictable accompaniment of end-stage renal disease (ESRD) in the absence of dietary phosphate restriction or supplemental phosphate binders. The consequences of hyperphosphatemia include the development and progression of secondary hyperparathyroidism and a predisposition to metastatic calcification when the product of serum calcium and phosphorus (Ca ؋ PO 4 ) is elevated. Both of these conditions may contribute to the substantial morbidity and mortality seen in patients with ESRD. We have analyzed the distribution of serum phosphorus in two large national, random, cross-sectional samples of hemodialysis patients who have been receiving dialysis for at least 1 year. Data were obtained from two special studies of the United States Renal Data System, the Case Mix Adequacy Study (1990) and the Dialysis Morbidity and Mortality Study Wave 1 (1993). The relative risk of death by serum phosphorus quintiles is described after adjusting for age at onset of ESRD, race, sex, smoking status, and the presence of diabetes, the acquired immunodeficiency syndrome, and/or neoplasm. Logistic regression analysis is then used to describe the demographic, comorbid, and laboratory parameters associated with high serum phosphorus. Serum phosphorus was similar in these two study populations and averaged 6.2 mg/dL. Ten percent of patients had levels greater than 9 mg/dL and at least 30% of each group had serum phosphorus levels greater than 7 mg/dL. The adjusted relative risk of death by serum phosphorus level was not uniform across all quintiles, being constant below a level of 6.5 mg/dL and increasing significantly above this level. The relative risk of death for those with a serum phosphorus greater than 6.5 mg/dL was 1.27 relative to those with a serum phosphorus of 2.4 to 6.5 mg/dL. This increased risk was not diminished by statistical adjustment for coexisting medical conditions, delivered dose of dialysis, nutritional parameters, or markers of noncompliance. Evaluation of predictors of serum phosphorus greater than 6.5 mg/dL revealed in multivariate analysis that younger age at onset of ESRD, female sex, white race, diabetes, active smoking, and higher serum creatinine levels were all significant predictors. Analysis of serum calcium revealed no correlation with relative risk of death. The Ca ؋ PO 4 product, however, showed a mortality risk trend similar to that seen with serum phosphorus alone. Those in the highest quintile of the Ca ؋ PO 4 product (G72 mg 2 /dL 2 ) had a relative mortality risk of 1.34 relative to those with products of 42 to 52 mg 2 /dL 2 . The relative mortality risk by log parathyroid hormone (PTH) level was elevated for patients with higher levels, but the mortality risk associated with hyperphosphatemia was independent of PTH. For hemodialysis patients who have been receiving dialysis for at least 1 year, we conclude that a large percentage have a serum phosphorus level above 6.5 mg/dL and that this places them at increased risk of death. This increased risk is independent of PTH. The mechanism(s) responsible for death is unknown, but may be related to an abnormally high Ca ؋ PO 4 product. Although mechanisms are not clearly established, this study supports the need for vigorous control of hyperphosphatemia to improve patient survival.

We evaluated the hierarchical risk groups for the estimated survival of WHO grade III glioma patients using recursive partitioning analysis (RPA). To our knowledge, this is the first study to address the results of RPA specifically for WHO grade III gliomas.

survival for patients with a single unresectable brain metastasis. WBRT and stereotactic radiosurgery should, therefore, be standard treatment for patients with a single unresectable brain metastasis and considered for patients with two or three brain metastases.

There has been an outbreak of the severe acute respiratory syndrome (SARS) worldwide. We report the clinical, laboratory, and radiologic features of 138 cases of suspected SARS during a hospital outbreak in Hong Kong. March 11 to 25, 2003, all patients with suspected SARS after exposure to an index patient or ward were admitted to the isolation wards of the Prince of Wales Hospital. Their demographic, clinical, laboratory, and radiologic characteristics were analyzed. Clinical end points included the need for intensive care and death. Univariate and multivariate analyses were performed.

Pregnancy and the puerperium are periods of increased risk for venous thromboembolism. 1 However, the absolute incidence of venous thromboembolism among pregnant or postpartum women is relatively low, ranging from 9 to 65 per 100,000 woman years for pulmonary embolism and deep vein thrombosis during pregnancy and from 70 to 180 per 100,000 woman years for pulmonary embolism and deep vein thrombosis during the puerperium. 2 Nevertheless, pulmonary embolism is still a major cause of death during pregnancy or the puerperium. 3 Universal heparin prophylaxis for all pregnant or postpartum women is not cost-effective because of the risk of heparin-induced throm-bocytopenia, bleeding, and osteoporotic fracture. 4, 5 Instead, heparin prophylaxis is recommended only for pregnant or postpartum women at high risk for venous thromboembolism, but there is disagreement regarding the characteristics that place such women at high risk. 4, 6 Although a history of venous thromboembolism may identify women at increased risk for recurrence during pregnancy or the puerperium, 4 most pregnancy-associated episodes of venous thromboembolism occur in the absence of such a history. 2, 7 Likewise, familial or acquired thrombophilia (eg, deficiency of antithrombin, protein C, or protein S or the presence of factor V R506Q [Leiden] mutation or a lupus anticoagulant-antiphospholipid antibody) clearly increases the risk, 8 but, again, most pregnancyassociated episodes of venous thromboembolism occur in the absence of a known thrombophilia. 2, 9 Screening for a familial or acquired thrombophilia among women contemplating pregnancy is not currently cost-effective. 10 Several other general patient characteristics (eg, age >35 years, weight >85 kg, varicose veins, or family history of venous thromboembolism) or pregnancy-related characteristics (eg, parity ≥4, preeclampsia, prolonged bed rest, preterm delivery, cesarean delivery, operative bleeding, or infection) have been suggested as risk factors, but the data supporting these suggestions are sparse and inconsistent.

A test for the ex ante efficiency of a given portfolio of assets is analyzed. The relevant statistic has a tractable small sample distribution. Its power function is derived and used to study the sensitivity of the test to the portfolio choice and to the number of assets used to determine the ex post mean-variance efficient frontier. Several intuitive interpretations of the test are provided, including a simple mean-standard deviation geometric explanation. A univariate test, equivalent to our multivariate-based method, is derived, and it suggests some useful diagnostic tools which may explain why the null hypothesis is rejected. Empirical examples suggest that the multivariate approach can lead to more appropriate conclusions than those based on traditional inference which relies on a set of dependent univariate statistics.

A simple risk score for prediction of contrast-induced nephropathy after This information is current as of OBJECTIVES We sought to develop a simple risk score of contrast-induced nephropathy (CIN) after percutaneous coronary intervention (PCI). BACKGROUND Although several risk factors for CIN have been identified, the cumulative risk rendered by their combination is unknown.

Activating mutations in the epidermal growth factor receptor gene (EGFR) confer hypersensitivity to the tyrosine kinase inhibitors gefitinib and erlotinib in patients with advanced non-small-cell lung cancer. We evaluated the feasibility of large-scale screening for EGFR mutations in such patients and analyzed the association between the mutations and the outcome of erlotinib treatment.

A generic preprocessing method for multivariate data, called orthogonal projections to latent structures (O-PLS), is described. O-PLS removes variation from X (descriptor variables) that is not correlated to Y (property variables, e.g. yield, cost or toxicity). In mathematical terms this is equivalent to removing systematic variation in X that is orthogonal to Y. In an earlier paper, Wold et al. (Chemometrics Intell. Lab. Syst. 1998; 44: 175±185) described orthogonal signal correction (OSC). In this paper a method with the same objective but with different means is described. The proposed O-PLS method analyzes the variation explained in each PLS component. The non-correlated systematic variation in X is removed, making interpretation of the resulting PLS model easier and with the additional benefit that the non-correlated variation itself can be analyzed further. As an example, near-infrared (NIR) reflectance spectra of wood chips were analyzed. Applying O-PLS resulted in reduced model complexity with preserved prediction ability, effective removal of noncorrelated variation in X and, not least, improved interpretational ability of both correlated and noncorrelated variation in the NIR spectra.