Disease relevance of Linear Models

• Multivariate linear regression analysis was used to determine if serum PSA levels were associated with the patient's age, race, or prostate volume in men without prostate cancer [1].
• Using multiple linear regression analysis to adjust for differences in age, obesity, alcohol consumption and regular exercise increased the differences in HDL cholesterol levels between smokers and nonsmokers [2].
• In nondiabetic Pimans, multivariate linear regressions of quantitative predictors of type 2 diabetes mellitus, including fasting plasma glucose, 2-h post-load plasma glucose, and body-mass index, showed significant inverse relations with IA when controlled for sex and age [3].
• The slope value of the linear regression line describing the relationship between pre-drive cycle length and pause duration for adenosine (15.8) was significantly greater than control (1.7), hypothermia (1.7), and cesium chloride (5.4) [4].
• In contracture due to oxygen and glucose deficiency, the linear regression had a slope of 6.06 (P less than 0.05) [5].

Psychiatry related information on Linear Models

• In a multiple stepwise linear regression analysis the following factors proved to increase healing incidence in decreasing order of importance: female sex, moderate alcohol consumption, abstinence from smoking, young age, and cimetidine treatment [6].
• The independent predictors of BMD were assessed through multiple linear regression, including age, height, weight, calcium intake, smoking, and physical activity [7].
• After adjustment for age, sex, income, social support, and duration of diabetes in linear regression models, there were significant graded relationships between greater depressive symptoms and higher serum levels of cholesterol and triglycerides (P<0.050) [8].
• MAIN OUTCOME MEASURES: The relationships between scores on the Brief, Repeatable Battery of Neuropsychological Tests in Multiple Sclerosis and 2 MRI measures (total lesion volume and brain to intracranial cavity volume ratio) were assessed using linear regression [9].
• Using multiple linear regression analysis, the blood oxygen level dependent (BOLD) response to smoking cues was correlated with severity of nicotine dependence assessed with the Fagerström Test of Nicotine Dependence (FTND) and with cue-induced craving [10].

High impact information on Linear Models

• The concentration of t-PA antigen correlated with that of the t-PA-PAI-1 complex in a linear regression model (squared correlation coefficient, 0.80; P<0.001) [11].
• Linear regression analysis revealed that nerve sorbitol content in the diabetics was inversely related to the number of myelinated fibers (P = 0.003) [12].
• After statistical adjustment for important prognostic factors (age and initial chronicity index) identified by multiple linear regression, the difference in the slopes between the group receiving prednisone and the group receiving cytotoxic drugs was significant (P less than 0.0001) [13].
• MAIN OUTCOME MEASURES: Proportion of days covered (PDC) by a statin in each quarter during the first year of therapy and every 6 months thereafter; predictors of suboptimal persistence during each interval (PDC <80%) were identified using generalized linear models for repeated measures [14].
• In linear regression analyses, the only significant predictors of change in CRP on a log scale were randomized pravastatin allocation and baseline CRP levels (P<.001 for both) [15].

Chemical compound and disease context of Linear Models

• Stepwise linear regression analysis showed that 60-degree MSV was strongly influenced by polyglutamine size and less by disease duration, whereas the reverse was found for ataxia score [16].
• We used multivariate linear regression to adjust for age, sex, total energy intake, body mass index, use of diuretics, beta-blockers, allopurinol, and uricosuric agents, self-reported hypertension and gout, serum creatinine level, and intake of alcohol [17].
• In normoinsulinemic subjects (fasting insulin < 100 pmol/L), dietary changes in postprandial triacylglycerols were significantly predicted by changes in phLPL, body weight, respiratory quotient (or fat oxidation), and the type of diet (stepwise multiple linear regression) [18].
• Multiple linear regression analysis revealed a strong correlation between myocardial norepinephrine depletion and left ventricular dysfunction, which was independent of all other variables [19].
• Multiple linear regression analyses showed significant associations between the severity score of atopic eczema and concentrations of allergen-specific IgE to dust mite (P =.032) and cat (P =.014) allergens after adjustment for sex, age, location, and parental predisposition [20].

Biological context of Linear Models

• Using multiple linear regression, both systolic and diastolic blood pressure remained significantly associated with darker skin color in the lower levels of socioeconomic status, independent of age, body mass index, and concentrations of blood glucose, serum urea nitrogen, serum uric acid, and urinary sodium and potassium [21].
• A linear model of the liver cell adhesion molecule L-CAM from embryonic chickens is proposed in terms of its orientation on the cell surface, the number, type, and distribution of carbohydrate moieties, and sites of phosphorylation [22].
• When the relationship between total cell CR3 and gestational age (n = 15) was analyzed, the correlation coefficient was .94 by linear regression, and the Spearman rank correlation was significant with P < .001 [23].
• Secretory rates (nmoles.min-1.-1 liver) for bile acids, cholesterol and phospholipids were determined and their mutual "linkage coefficients" (nmoles of secretory increment per 1 nmole of bile acid secreted) calculated by linear regression analysis [24].
• Logistic and linear regression analyses were performed to determine the association of nitrate use with new radiographic findings of hip OA, adjusting for age, weight, height, bone mineral density, and estrogen [25].

Anatomical context of Linear Models

• When the data were analyzed by multiple linear regression, the percentage of CD4+ cells had a positive effect on these two parameters in patients without AIDS, whereas the percentage of CD4+ cells had no significant effect on unstimulated and IL-2-stimulated NK cell activity in these patients [26].
• The T cell subpopulation was 5'NT deficient compared to the normal controls, shown as a significant linear regression of T cell 5'NT on the LI% [27].
• These simulated curves were used to construct the statistical linear model and to test voxel-by-voxel in healthy subjects the hypothesis that dopamine is released in the ventral striatum during periods of unexpected monetary gains, but not during periods of unexpected monetary loss [28].
• Multivariate linear regression models examined independent effects of diabetes, lean mass, fat mass, visceral fat, and fasting insulin on BMD and bone volume while adjusting for relevant covariates [29].
• MAIN MEASUREMENTS: Changes in CD4 lymphocyte count and CD4 percentage over time were analysed using linear regression methods for the 271 white and 69 black participants who had at least four semiannual CD4 lymphocyte measurements [30].

Associations of Linear Models with chemical compounds

• By linear regression analysis of the correlation (r = 0.99) the number of heparin molecules/particle required to promote total inactivation of bound C3b by purified control proteins was 13.8 X 10(6) [31].
• Linear regression analyses revealed that plasma triglyceride concentrations in all fed, treated rats correlated significantly with triglyceride entry but not very uniformly with plasma or tissue LPL activity [32].
• Variability across the studies in the proportion of tumors expressing MDR1/gp170 was assessed by use of chi-squared tests of homogeneity, weighted means, and weighted linear regression [33].
• Linear regression analysis was used to analyze correlations between the various protein levels, DNA-PK activity, or Rad51 foci formation and melphalan cytotoxicity [34].
• In multivariate log-linear models virtually all the treatment effect was due to zidovudine [35].

Gene context of Linear Models

• To test this hypothesis, we used step-wise multiple linear regression to examine 10 CAG-containing genes for possible influences on SCA2 age at onset [36].
• Using a log-linear model, differences in ER, PR, and HER-2 staining were all in the direction of stronger staining in the cores, and were statistically significant [37].
• Linear regression analysis found no association between NAT2 activity and the levels (unmetabolized plus acid-labile conjugates) of MeIQx or PhIP excreted in urine [38].
• Linear regression analyses showed that lower total MeIQx (unmetabolized plus the N2-glucuronide and sulfamate metabolites) in urine was associated with higher CYP1A2 activity, whereas total PhIP (unmetabolized plus conjugated) in urine showed no association to CYP1A2 activity [38].
• Using hierarchical linear regression, a significant independent association was found between the DAT1 10/10-repeat genotype and measures of selective attention and response inhibition after adjusting for age, IQ, and ADHD symptoms [39].

Analytical, diagnostic and therapeutic context of Linear Models

• The temporal profile of the effect site concentration of remifentanil, estimated from a pharmacokinetic model, was incorporated into a linear model of the fMRI data [40].
• Multiple linear regression analysis confirmed that the degree of perfusion of the infarct artery (partial r = 0.58, p = 0.001) was a more important predictor of volume change than was infarct size measured by peak creatine kinase (partial r = 0.30, p = 0.009) or QRS score (partial r = 0.20, p = 0.087) [41].
• BCM was modestly underestimated by BIA compared with the dual-energy x-ray absorptiometry/NaBr standard and was adjusted by linear regression [42].
• Using multiple linear regression, GH status at retesting is predicted by BED, age at irradiation, and use of chemotherapy [43].
• The PC3 tumor volume growth curves were log transformed and fitted using linear regression [44].

References