SEARCH · PubMed Health

Results for “Linear Models”

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 1,063 records · Page 59Linked to original sources

PubMed1998

The linear-quadratic model and most other common radiobiological models result in similar predictions of time-dose relationships.

One of the fundamental tools in radiation biology is a formalism describing time-dose relationships. For example, there is a need for reliable predictions of radiotherapeutic isoeffect doses when the temporal exposure pattern is changed. The most commonly used tool is now the linear-quadratic (LQ) formalism, which describes fractionation and dose-protraction effects through a particular functional form, the generalized Lea-Catcheside time factor, G. We investigate the relationship of the LQ formalism to those describing other commonly discussed radiobiological models in terms of their predicted time-dose relationships. We show that a broad range of radiobiological models are described by formalisms in which a perturbation calculation produces the standard LQ relationship for dose fractionation/protraction, including the same generalized time factor, G. This approximate equivalence holds not only for the formalisms describing binary misrepair models, which are conceptually similar to LQ, but also for formalisms describing models embodying a very different explanation for time-dose effects, namely saturation of repair capacity. In terms of applications to radiotherapy, we show that a typical saturable repair formalism predicts practically the same dependences for protraction effects as does the LQ formalism, at clinically relevant doses per fraction. For low-dose-rate exposure, the same equivalence between predictions holds for early-responding end points such as tumor control, but less so for late-responding end points. Overall, use of the LQ formalism to predict dose-time relationships is a notably robust procedure, depending less than previously thought on knowledge of detailed biophysical mechanisms, since various conceptually different biophysical models lead, in a reasonable approximation, to the LQ relationship including the standard form of the generalized time factor, G.

Brachytherapy↗

PubMed1995

Capture-recapture and multiple-record systems estimation I: History and theoretical development. International Working Group for Disease Monitoring and Forecasting.

This paper reviews the historical background and the theoretical development of models for the analysis of data from capture-recapture or multiple-record systems for estimating the size of closed populations. The models and methods were originally developed for use in fisheries and wildlife biology and were later adapted for use in connection with human populations. Application to epidemiology came much later. The simplest capture-recapture model involves two lists or samples and has four key assumptions: that the population is closed, that individuals can be matched from capture to recapture, that capture in the second sample is independent of capture in the first sample, and that the capture probabilities are homogeneous across all individuals in the population. Log-linear models provide a convenient representation for this basic capture-recapture model and its extensions to K lists. The paper provides an overview for these models and illustrates how they allow for dependency among the lists and heterogeneity in the population. The use of log-linear models for estimation in the presence of both dependence and heterogeneity is illustrated on a four-list example involving ascertainment of diabetes using data gathered in 1988 from residents of Casale Monferrato, Italy. The final section of the paper discusses techniques for model selection in the context of models for estimating the size of populations.

Bias↗

PubMed2002

Discovering subpopulation structure with latent class mixed models.

The linear mixed model is a well-known method for incorporating heterogeneity (for example, subject-to-subject variation) into a statistical analysis for continuous responses. However heterogeneity cannot always be fully captured by the usual assumptions of normally distributed random effects. Latent class mixed models offer a way of incorporating additional heterogeneity which can be used to uncover distinct subpopulations, to incorporate correlated non-normally distributed outcomes and to classify individuals. The methodology is motivated with examples in health care studies and a detailed illustration is drawn from the Nutritional Prevention of Cancer trials. Latent class models are used with longitudinal data on prostate specific antigen (PSA) as well as incidence of prostate cancer. The models are extended to accommodate prostate cancer as a survival endpoint; this is compared to treating it as a binary endpoint. Four subpopulations are identified which differ both with regard to their PSA trajectories and their incidence rates of prostate cancer.

Adolescent↗

PubMed2006

Ambulatory pulse pressure and progression of urinary albumin excretion in older patients with type 2 diabetes mellitus.

We studied whether ambulatory blood pressure monitoring added to office blood pressure in predicting progression of urine albumin excretion over 2 years of follow-up in a multiethnic cohort of older people with type-2 diabetes mellitus. Participants in the Informatics for Diabetes Education and Telemedicine study underwent a baseline evaluation that included office and 24-hour ambulatory blood pressure measurement and a spot urine measurement of albumin-to-creatinine ratio (ACR). Measurements of albumin-to-creatinine ratio were repeated 1 and 2 years later. In bivariate analyses, ambulatory 24-hour pulse pressure was the blood pressure variable most strongly associated with follow-up ACR. Repeated-measures mixed linear models (n = 1040) were built adjusting for baseline ACR ratio, clustered randomization, time to follow-up, and multiple covariates. When both were entered into the model, ambulatory 24-hour pulse pressure and office pulse pressure were independently associated with follow-up ACR (beta [SE] = 0.010 [0.002], P < 0.001, and 0.004 [0.001], P = 0.002, respectively). Cox proportional hazards models examined associations with progression of albuminuria in 954 participants without macroalbuminuria at baseline, adjusting for all of the covariates independently associated with follow-up ACR in mixed linear models. Ambulatory 24-hour pulse pressure, but not office pulse pressure, was independently associated with progression of albuminuria (P = 0.015 and 0.052, respectively). The adjusted hazards ratio (95% CI) per each 10-mm Hg increment in ambulatory pulse pressure was 1.23 (1.04 to 1.42). In conclusion, ambulatory pulse pressure may provide additional information to predict progression of albuminuria in elderly diabetic subjects above and beyond office blood pressure.

Aged↗

PubMed1996

Plasma pharmacokinetics, bioavailability, and tissue distribution in CD2F1 mice of halomon, an antitumor halogenated monoterpene isolated from the red algae Portieria hornemannii.

The purpose of the present study was to define the plasma pharmacokinetics, bioavailability, and tissue distribution in mice of halomon, a halogenated monoterpene from Portieria hornemanii that is active in vitro against brain-, renal-, and colon-cancer cell lines. Halomon formulated in cremophor:ethanol:0.154 M NaCl (1:1:6, by vol.) was injected i.v. at 20, 60, 90, or 135 mg/kg into female CD2F1 mice. Doses of 135 mg/kg were also given i.p., s.c., and by enteral gavage to female CD2F1 mice and i.v. to male CD2F1 mice. Plasma halomon concentrations were measured with a gas-chromatography system using electron-capture detection. Halomon concentrations were also determined in the brains, hearts, lungs, livers, kidneys, spleens, skeletal muscles, fat, red blood cells, and, if present, testes of mice given 135 mg/kg i.v. Halomon plasma pharmacokinetics were well fit by a two-compartment, open linear model and were linear between 20 and 135 mg/kg. Population estimates of parameters describing halomon plasma pharmacokinetics in female CD2F1 mice were developed with a standard two-stage technique and also by simultaneous modeling of data from 20-, 60-, 90-, and 135-mg/kg i.v. studies in female mice. Halomon bioavailability was 45%, 47%, and 4% after i.p., s.c., and enteral dosing, respectively. Urinary excretion of the parent compound was minimal. Halomon was distributed widely to all tissues studied but was concentrated and persisted in fat. Halomon concentrations measured in the brain were comparable with concomitant concentrations detected in plasma and most other tissues. These data and models are helpful in the simulation and evaluation of conditions produced by preclinical screening and toxicology studies.

Administration, Oral↗

PubMed2002

Erosions in inflammatory polyarthritis are symmetrical regardless of rheumatoid factor status: results from a primary care-based inception cohort of patients.

BACKGROUND AND AIMS: Symmetry is considered an important criterion for the differentiation of rheumatoid arthritis (RA) from other forms of inflammatory polyarthritis (IP), particularly those that are seronegative. Because of the inclusion of symmetry in the diagnostic and classification process, however, its true occurrence in RA cannot be assessed. As a surrogate, peripheral inflammatory arthropathies associated with rheumatoid factor production may be more likely to be symmetrical. We examined the degree of symmetry of erosions in an unselected cohort of patients with IP and tested the hypothesis that the presence of rheumatoid factor (RF) is associated with greater symmetry. METHODS: All patients registered with The Norfolk Arthritis Register (NOAR; a UK primary-care based cohort of patients with IP with annual follow-up) and who had radiographs performed at the fifth anniversary from notification were included in the analysis. Radiographs of the hands and feet were read using the Larsen method; a score of 2 or more in any particular joint indicated an erosion. Log-linear modelling was used to determine the symmetry of erosions between right and left for the following joint groups: wrists, metacarpophalangeal joints, proximal interphalangeal joints and metatarsophalangeal joints. Log-linear modelling was also used to determine the influence of RF on symmetry. RESULTS: Five hundred and thirty-seven patients contributed to the analysis. The median time to performing radiographs was 69 months (interquartile range 65.5-74.8) from the onset of symptoms. A total of 212 (39%) patients had erosive disease. Overall, IP was found to be a symmetrical disease. Despite there being more erosions in RF-positive patients, there was no greater excess of symmetry in RF-positive compared with RF-negative patients. CONCLUSION: Radiographically, IP is a symmetrical disease irrespective of RF status. The use of symmetry as an important feature in identifying subgroups of patients with IP, such as RA, is challenged.

Arthritis, Rheumatoid↗

PubMed2002

Naevi in allogeneic bone marrow transplantation recipients: the effect of graft-versus-host disease on naevi.

BACKGROUND: Non-melanoma skin carcinoma is more common in transplant recipients, probably because of immunosuppression. An increased risk of developing melanoma could be a late effect of transplantation. The number of naevi, which is a risk marker for melanoma, is increased in renal transplant recipients of all ages and may be related to immunosuppression. The risk of melanoma has been suspected to be particularly high after bone marrow transplantation. Cutaneous graft-versus-host disease (GVHD) might be an interesting model for the study of interactions between naevi and the immune system. OBJECTIVE: To assess whether aBMT exposes an individual to a particularly high risk of melanoma, using naevi as a surrogate measure of the risk. To improve our knowledge of the effect of the immune system on naevi, using GVHD as a model. METHODS: We carried out an epidemiological case-control study with two age-, sex- and hair colour-matched controls for each case. The results were analysed with analysis of variance, a general linear model analysis and multivariate analysis. RESULTS: The number of naevi was not significantly increased in aBMT patients, as compared with controls, although there was a significant excess on the palms and legs. In exploratory subgroup case-control comparisons and with the general linear model, patients who were conditioned with a combination of two alkylating drugs at high doses, and patients who had an aBMT before the age of 20 years tended to have a higher count of naevi (P = 0.002 and P = 0.06, respectively). Conversely, there was a trend in favour of a lower count of naevi in patients with diffuse skin lesions of chronic GVHD (P = 0.01). These data were corroborated by multivariate analysis, which showed that conditioning with high-dose chemotherapy, the absence of severe chronic cutaneous GVHD and a young age at transplantation were the main variables that independently predicted an excess of naevi. CONCLUSIONS: This study of aBMT patients confirms that chemotherapy stimulates naevus growth. It also suggests for the first time that diffuse lesions of chronic cutaneous GVHD are associated with a decreased number of naevi. Whether allo-immunity, chronic skin inflammation or the masking of naevi by pigmentation and fibrosis is responsible for the decreased number of naevi requires further investigation. With respect to the long-term risk of melanoma in aBMT recipients, our results support an increased risk particularly when aBMT is performed at a young age, and when conditioning is with high doses of alkylating drugs.

Adolescent↗

PubMed1989

A conjecture on the interface mechanics in fractures based on the interpretation of impulse vests.

Time-domain tests for monitoring fracture healing, undertaken by the authors and workers elsewhere, show a repeatable characteristic signal behaviour which is difficult to explain using the linear models prevailing in biomechanics. The paper proposes a simple non-linear model which, it is believed, will reproduce the observed signal behaviour and is consistent with the biphasic stiffness model previously identified in static mechanical testing for monitoring of fracture healing. This, in turn, correlates with the two stages of callus formation in the accepted biological model.

Biomechanical Phenomena↗

PubMed1982

Covariance analysis for case-control studies with small blocks.

Medical and epidemiological studies often employ a blocked design in order to control for some types of variation. For example, environmental and, to some extent, genetic factors may be controlled by blocking on sibs. The most common situation is one in which a subject in one group is blocked with one or more subjects in another. Additional factors may be considered as covariates in the analysis. Conditional likelihood methods give rise to consistent estimators of the parameters of the log-linear model. These estimators are generalized, and cases where there are small numbers of subjects in each group or where there are more than two groups are discussed. The resulting models can be considered as special cases of the general-linear model, and therefore the analyses can often be carried out with existing computer programs. The methods are illustrated by use of data from a case-control study in which each block contains one case and four controls.

Adolescent↗

PubMed2006

Estimating average annual percent change for disease rates without assuming constant change.

The annual percent change (APC) is often used to measure trends in disease and mortality rates, and a common estimator of this parameter uses a linear model on the log of the age-standardized rates. Under the assumption of linearity on the log scale, which is equivalent to a constant change assumption, APC can be equivalently defined in three ways as transformations of either (1) the slope of the line that runs through the log of each rate, (2) the ratio of the last rate to the first rate in the series, or (3) the geometric mean of the proportional changes in the rates over the series. When the constant change assumption fails then the first definition cannot be applied as is, while the second and third definitions unambiguously define the same parameter regardless of whether the assumption holds. We call this parameter the percent change annualized (PCA) and propose two new estimators of it. The first, the two-point estimator, uses only the first and last rates, assuming nothing about the rates in between. This estimator requires fewer assumptions and is asymptotically unbiased as the size of the population gets large, but has more variability since it uses no information from the middle rates. The second estimator is an adaptive one and equals the linear model estimator with a high probability when the rates are not significantly different from linear on the log scale, but includes fewer points if there are significant departures from that linearity. For the two-point estimator we can use confidence intervals previously developed for ratios of directly standardized rates. For the adaptive estimator, we show through simulation that the bootstrap confidence intervals give appropriate coverage.

Biometry↗

PubMed2004

Lung cancer mortality predictions for Ireland 2001-2015 and current trends in North Western Europe.

UNLABELLED: PURPOSE AND METHODOLOGY: Poisson regression models of Irish lung cancer mortality data for the years 1954-2000 were used to predict European age standardized mortality rates (EASMRs) and numbers of deaths for the period 2001-2015. Predictions for the whole population and for the age group 0-64 years were the main focus of this analysis. The models employed in this analysis were written for the STATA 6.0 STATISTICAL PACKAGE FOR WINDOWS. Linear models were applied, in general, to upward trends and log-linear models to downward trends. RESULTS: For males of all ages, EASMRs are expected to fall significantly from 59.7 per 100,000 person years (100,000 PY(-1)) in 2000 to 47.2 in the period 2011-2015. Female EASMRs are expected to fall from 28.3 100,000 PY(-1) in 2000 to 27.8 in 2011-2015. The EASMR for males under 65 years of age is predicted to continue its decline from 20.6 100,000 PY(-1) in 2000 to 11.7 in 2011-2015. The rate for females in this same age group is predicted to decline from 10.7 100,000 PY(-1) in 2000 to 7.0 in 2001-2015. Recent lung cancer trends In Ireland were also compared to other countries in North Western Europe.

Adolescent↗

PubMed2006

Empirical array quality weights in the analysis of microarray data.

BACKGROUND: Assessment of array quality is an essential step in the analysis of data from microarray experiments. Once detected, less reliable arrays are typically excluded or "filtered" from further analysis to avoid misleading results. RESULTS: In this article, a graduated approach to array quality is considered based on empirical reproducibility of the gene expression measures from replicate arrays. Weights are assigned to each microarray by fitting a heteroscedastic linear model with shared array variance terms. A novel gene-by-gene update algorithm is used to efficiently estimate the array variances. The inverse variances are used as weights in the linear model analysis to identify differentially expressed genes. The method successfully assigns lower weights to less reproducible arrays from different experiments. Down-weighting the observations from suspect arrays increases the power to detect differential expression. In smaller experiments, this approach outperforms the usual method of filtering the data. The method is available in the limma software package which is implemented in the R software environment. CONCLUSION: This method complements existing normalisation and spot quality procedures, and allows poorer quality arrays, which would otherwise be discarded, to be included in an analysis. It is applicable to microarray data from experiments with some level of replication.

Algorithms↗

PubMed2006

A material sensitivity study on the accuracy of deformable organ registration using linear biomechanical models.

Model-based deformable organ registration techniques using the finite element method (FEM) have recently been investigated intensively and applied to image-guided adaptive radiotherapy (IGART). These techniques assume that human organs are linearly elastic material, and their mechanical properties are predetermined. Unfortunately, the accurate measurement of the tissue material properties is challenging and the properties usually vary between patients. A common issue is therefore the achievable accuracy of the calculation due to the limited access to tissue elastic material constants. In this study, we performed a systematic investigation on this subject based on tissue biomechanics and computer simulations to establish the relationships between achievable registration accuracy and tissue mechanical and organ geometrical properties. Primarily we focused on image registration for three organs: rectal wall, bladder wall, and prostate. The tissue anisotropy due to orientation preference in tissue fiber alignment is captured by using an orthotropic or a transversely isotropic elastic model. First we developed biomechanical models for the rectal wall, bladder wall, and prostate using simplified geometries and investigated the effect of varying material parameters on the resulting organ deformation. Then computer models based on patient image data were constructed, and image registrations were performed. The sensitivity of registration errors was studied by perturbating the tissue material properties from their mean values while fixing the boundary conditions. The simulation results demonstrated that registration error for a subvolume increases as its distance from the boundary increases. Also, a variable associated with material stability was found to be a dominant factor in registration accuracy in the context of material uncertainty. For hollow thin organs such as rectal walls and bladder walls, the registration errors are limited. Given 30% in material uncertainty, the registration error is limited to within 1.3 mm. For a solid organ such as the prostate, the registration errors are much larger. Given 30% in material uncertainty, the registration error can reach 4.5 mm. However, the registration error distribution for prostates shows that most of the subvolumes have a much smaller registration error. A deformable organ registration technique that uses FEM is a good candidate in IGART if the mean material parameters are available.

Biomechanical Phenomena↗

PubMed1999

The "thermolabile" variant of methylenetetrahydrofolate reductase and neural tube defects: An evaluation of genetic risk and the relative importance of the genotypes of the embryo and the mother.

Recent reports have implicated the "thermolabile" (T) variant of methylenetetrahydrofolate reductase (MTHFR) in the causation of folate-dependent neural tube defects (NTDs). We report herein the largest genetic study of NTD cases (n=271) and families (n=218) to date, establishing that, in Ireland, the "TT" genotype is found in 18.8% of cases versus 8.3% of controls (odds ratio 2.57; confidence interval [CI] 1.48-4.45; P=.0005). The maternal and paternal TT genotypes have intermediate frequencies of 13.8% and 11.9%, respectively, indicating that the predominant MTHFR-related genetic effect acts via the TT genotype of the developing embryo. Analysis of the 218 family triads of mother, father, and affected child with log-linear models supports this interpretation, providing significant evidence that the case TT genotype is associated with NTDs (P=.02) but no evidence of a maternal TT genotypic effect (P=. 83). The log-linear model predicted that the risk of NTDs conferred by the case TT genotype is 1.61 (CI 1.06-2.46), consistent with the paramount importance of the case TT genotype in determining risk. There is no compelling evidence for more than a modest additional risk conferred by a maternal TT genotype. These results favor a biological model of MTHFR-related NTD pathogenesis in which suboptimal maternal folate status imposes biochemical stress on the developing embryo, a stress it is ill-equipped to tolerate if it has a TT genotype.

Alleles↗

PubMed1996

Explained variation for logistic regression.

Different measures of the proportion of variation in a dependent variable explained by covariates are reported by different standard programs for logistic regression. We review twelve measures that have been suggested or might be useful to measure explained variation in logistic regression models. The definitions and properties of these measures are discussed and their performance is compared in an empirical study. Two of the measures (squared Pearson correlation between the binary outcome and the predictor, and the proportional reduction of squared Pearson residuals by the use of covariates) give almost identical results, agree very well with the multiple R2 of the general linear model, have an intuitively clear interpretation and perform satisfactorily in our study. For all measures the explained variation for the given sample and also the one expected in future samples can be obtained easily. For small samples an adjustment analogous to Radj2 in the general linear model is suggested. We discuss some aspects of application and recommend the routine use of a suitable measure of explained variation for logistic models.

Likelihood Functions↗

PubMed1989

Models for genetic analysis of dystocia and calf mortality.

Calvings of 106,750 Israeli Holstein heifers were analyzed for dystocia and calf mortality, which were scored dichotomously. Models fitted included herd-year-season, sex of calf, calving age, calving month, sire of cow, sire of calf, and groups of sire of cow and calf. Herd-year-season, sire of cow and of calf, and residuals were random with diagonal variance-covariance matrices. Traits were analyzed with and without a herd-year-season effect, and dystocia was analyzed separately for male and female calvings and with both sexes combined. Variance components were estimated by REML for linear models and by the counterpart of REML for threshold models. Heritability estimates were two to four times larger in threshold than in linear models, but correlations between corresponding sire evaluations were all greater than .9. Correlations between evaluations computed with and without herd-year-season effects were .9 for sire of calf evaluations for dystocia and greater than .97 for all other evaluations. Correlations between sire evaluations computed separately for male and female calvings were between .3 and .7. Thus, calving difficulty is expressed differently in male and female calvings. Genetic trends were unfavorable for dystocia but favorable for calf mortality. Phenotypic trends were curvilinear with maxima in 1982 and 1983 for calf mortality and dystocia.

Age Factors↗

PubMed2001

Examination of temporal trends in the incidence of childhood leukaemias and lymphomas provides aetiological clues.

The age-sex distributions and temporal trends in incidence of leukaemia and lymphoma from the Manchester Children's Tumour Registry (MCTR), 1954-1998, are reported. This 45-year study includes 1795 children, all of whom had a histologically and/or cytologically verified leukaemia or lymphoma. At the time of their diagnoses all the children were under 15 years of age and were resident in a geographically defined area of northwest England covered by the MCTR. Log-linear modelling identified significant linear increases in acute lymphoblastic leukaemia (ALL) (average annual increase 0.7%; P= 0.005) and in Hodgkin's disease (HD) (1.2%, P=0.04), but not in acute myeloid leukaemia (AML), nor in non-Hodgkin's lymphoma (NHL). The increase in ALL was most pronounced amongst males, aged 1-4 years, and is likely to be due to precursor B-cell leukaemias. The increases in ALL and HD are discussed in relation to current hypotheses suggesting a role for infection. Additionally, a non-linear cohort effect was identified for NHL (P= 0.008), which may indicate the involvement of environmental factors other than infection.

Acute Disease↗

PubMed1975

Best linear unbiased estimation and prediction under a selection model.

Mixed linear models are assumed in most animal breeding applications. Convenient methods for computing BLUE of the estimable linear functions of the fixed elements of the model and for computing best linear unbiased predictions of the random elements of the model have been available. Most data available to animal breeders, however, do not meet the usual requirements of random sampling, the problem being that the data arise either from selection experiments or from breeders' herds which are undergoing selection. Consequently, the usual methods are likely to yield biased estimates and predictions. Methods for dealing with such data are presented in this paper.

Animals↗