PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Regression Analysis”

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

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

At least 901 records · Page 50Linked to original sources

Factors affecting the decision to exit dairy farming: a two-stage regression analysis.

Substantial attention has focused on the dairy industry because of a concern that many producers are getting out of dairying. Although low milk prices are postulated as a primary reason for exits from dairying, other factors may be important as well. Data from a representative 64-farm subset of a 2002 survey of dairy producers in Maine were used in the current study. Of the 64 farms, 15 indicated an imminent exit from dairying, whereas 49 dairy farms expected to remain in business for > or =5 yr. A binary choice logit regression model, based upon the dependent variable decision to exit or remain in the industry, was used as part of a 2-stage regression process to ascertain why dairy producers are choosing to leave the industry. The hypothesis states that the decision is a function of 3 independent variable categories: demographic, efficiency, and opportunity costs. Four variables were revealed that significantly influence the exit decision. Older producers, higher off-farm income, lower returns over variable cost, and greater diversification of farm income were more likely associated with a decision to leave dairy farming. Because factors other than milk price are involved in exit decisions, perhaps national or regional dairy programs should consider strategies beyond price supports to provide for a stable dairy industry and a reduction in the rate of dairy farm exits.

Animals↗

[Regression analysis of the heart rate increase in children during linearly increasing work load on a bicycle ergometer].

E. A. Müller (1950) has introduced a procedure for the measurement of the "Leistungspulsindex" (LPI) based on the determination of a linear regression between heart rate increase and work load. For this purpose the work load on a bicycle ergometer is increased in proportion to time (about 10 W/min) and the heart rate is recorded continuously. We have performed measurements in children (10--11 years of age) and found that the above mentioned regression is markedly nonlinear and can be described best by two overlapping linear regression lines with different slopes (b11 and b12). We propose that under these conditions the regression coefficient b11 of the first part of the relation can be assumed as LPI. In our analysis we found the linear part of the heart rate to increase between the end of min 2 and min 6 (+/- 1.4) of the experiment in boys. In girls we found the linear part of the regression only until min 3.8 (+/- 1.4). One consequence of our results is related to the calculation of the PWC170 by linear extrapolation which appears to be impossible in all cases in which the relation between work load and heart rate is nonlinear. It seems possible to relate that part of the heart rate increase where the slope suddenly changes to the endurance limit.

Child↗

Conditional regression analysis of the exposure-disease odds ratio using known probability-of-exposure values.

Conditional inference methods are proposed for the odds ratio between binary exposure and disease variables when only the probability of exposure is known for each study subject. We develop a conditional likelihood approach that removes nuisance parameters and permits inferences to be made about important parameters in log odds ratio regression models. We also discuss a heuristic procedure based on estimating the (unknown) number of truly exposed individuals; this procedure provides a simple framework for interpreting our likelihood-based statistics, and leads to a Mantel-Haenszel-type estimator and a goodness-of-fit test. As an example of the use of this methodology, we present an analysis of some genetic data of Swift et al. (1976, Cancer Research 36, 209-215).

Adult↗

Crossed cerebellar diaschisis in patients with cortical infarction: logistic regression analysis to control for confounding effects.

BACKGROUND AND PURPOSE: Crossed cerebellar diaschisis (CCD) refers to reduced metabolism and blood flow in the cerebellar hemisphere contralateral to a cerebral lesion. Many cortical areas have been reported to cause CCD without consideration of confounding factors. We performed single-photon emission computed tomography (SPECT) in patients with cortical infarction to identify regions independently related to CCD, controlling for possible confounding effects. METHODS: Patients with unilateral cortical infarction (n=113; 75 male, 38 female; mean+/-SD age, 66+/-13 years) underwent SPECT of the brain with N-isopropyl-p-[(123)I]iodoamphetamine ((123)I-IMP). Regional cerebral blood flow was measured autoradiographically. Asymmetry indices (AIs) were calculated on the basis of ratios representing symmetrical regional cerebral blood flow in the cerebellum and 16 cerebral regions. CCD was defined as AI for cerebellum >0.1. AIs for 16 cortical regions were considered for both dichotomous and continuous variables for analysis of CCD occurrence by means of backward logistic regression. RESULTS: For dichotomized variables, hypoperfusion of postcentral (odds ratio [OR]=7.607; 95% CI, 2.299 to 25.174) and supramarginal (OR=3.916; 95% CI, 1.394 to 11.003) regions independently influenced CCD. For continuous variables, hypoperfusion of postcentral (OR=1.044; 95% CI, 1.019 to 1.068) and supramarginal (OR=1.021; 95% CI, 1.001 to 1.041) regions (and, as a negative factor, medial occipital regions; OR=0.942; 95% CI, 0.895 to 0.991) independently influenced CCD. CONCLUSIONS: Many cortical areas apparently do not contribute to CCD. Correspondence of CCD between dichotomized and continuous analyses suggests that location of a lesion, not severity, is the main determinant of CCD.

Adult↗

The interaction of eltanolone and fentanyl with special reference to logistic regression analysis.

UNLABELLED: We investigated whether fentanyl decreases the serum concentrations of the steroid anesthetic eltanolone effective in producing loss of consciousness in 50% of patients (EC50induction) and in preventing movement at skin incision in 50% of patients (EC50incision). For anesthetic induction, patients received effect-site target concentrations of fentanyl 0.0, 1.5, 3.0, or 4.5 ng/mL and eltanolone 500, 750, 1000, or 1200 ng/mL. Loss of response to verbal command was assessed after 10 min. For incision, patients received effect-site target concentrations of fentanyl 0.5,1.5, 3.0, or 4.5 ng/mL and eltanolone 547-2926 ng/mL. Movement at incision was assessed at least 10 min after new targets were entered. Probability of loss of consciousness and of movement versus arterial serum concentration combinations were analyzed by logistic regression. Dixon up-down analysis was used to estimate ET50incision effective target concentration combinations. In the absence of fentanyl, anesthesia was induced in only 1 of 12 patients, which suggests that the EC50induction is >1500 ng/mL at fentanyl 0.0 ng/mL. With fentanyl (38 patients), eltanolone EC50induction was independent of fentanyl concentration, calculated as 628 ng/mL. For the incision phase (52 patients), logistic regression failed to generate a valid model. Dixon analysis (43 patients) produced an eltanolone ET50incision of 2288 ng/mL at fentanyl targets of 0.5 ng/mL, 754 ng/mL at 1.5 ng/mL, 735 ng/mL at 3.0 ng/mL, and 645 ng/mL at 4.5 ng/mL. Fentanyl reduced the serum concentration of eltanolone required to produce loss of consciousness and the target concentration of eltanolone required to prevent movement to skin incision. IMPLICATIONS: Fentanyl reduced the serum concentration of eltanolone required to produce loss of consciousness and the target concentration of eltanolone required to prevent movement to skin incision. Future interaction studies of this nature using logistic regression should model responses to hypnotic alone separately from responses to hypnotic-analgesic combinations.

Adolescent↗

Multiple regression analysis of dental status and related food behaviour of French Canadian adolescents.

Variables included in a theoretical model of the caries process have been studied in an adolescent population. Four multiple regression analyses were performed to uncover significant explanatory factors for: DMF(T) index, plaque index, frequency of sweets consumption between meals and frequency of sweets consumption at meals. Each analysis produced a significant R2 explaining respectively 29%, 15%, 74% and 62% of the variance. The results demonstrated the contribution of the different factors included in the proposed model. They suggested that plaque index was by far the most significant factor of the DMF(T) index. Sweets consumption was a significant determinant of caries experience independent of any relation to plaque index. Adolescent girls, even with less plaque accumulation, had a higher caries index than boys of the same age. Parent's education was related to dental status of children at time of adolescence.

Adolescent↗

A regression analysis of test bias on the WISC-R for Anglos and Chicanos referred for psychological services.

Predictive validity of the Wechsler Intelligence Scale for Children-Revised was investigated across race for 174 Chicano and 94 Anglo children referred for psychological services. Regression lines for the prediction of achievement were compared across race through the Potthoff analysis, which provides a simultaneous test of slope and intercept values. Results of these comparisons generally supported the predictive validity of the WISC-R across race with this referral sample of young children.

Child↗

Stepwise selection in small data sets: a simulation study of bias in logistic regression analysis.

Stepwise selection methods are widely applied to identify covariables for inclusion in regression models. One of the problems of stepwise selection is biased estimation of the regression coefficients. We illustrate this "selection bias" with logistic regression in the GUSTO-I trial (40,830 patients with an acute myocardial infarction). Random samples were drawn that included 3, 5, 10, 20, or 40 events per variable (EPV). Backward stepwise selection was applied in models containing 8 or 16 pre-specified predictors of 30-day mortality. We found a considerable overestimation of regression coefficients of selected covariables. The selection bias decreased with increasing EPV. For EPV 3, 10, or 40, the bias exceeded 25% for 7, 3, and 1 in the 8-predictor model respectively, when a conventional selection criterion was used (alpha = 0.05). For these EPV values, the bias was less than 20% for all covariables when no selection was applied. We conclude that stepwise selection may result in a substantial bias of estimated regression coefficients.

Aged↗

[Cluster and multiple regression analysis of leptospirosis epidemic factors].

Ten-year surveillance of Rice-field-typed leptospirosis was carried out in high-infected foci. Twenty two parameters of epidemic data were divided into 3 categories and cluster-analyzed. 5 typical variables were selected for multiple regression equation. The equation contained four factors, i.e. rodents density, carriage rates of main animal hosts, GMT of population antibody against leptospira and quantity of rainfall in August. The expected morbidity of leptospirosis calculated by the equation were roughly identical to the real morbidity.

Animals↗

Prevalence of anal incontinence according to age and gender: a systematic review and meta-regression analysis.

CONTEXT: Anal incontinence is increasingly being recognised as a significant cause of physical and psychological morbidity with implications for healthcare provision within the community. There is controversy about which population groups are most disadvantaged by this chronic condition. OBJECTIVES: The aim of this study was to evaluate the prevalence of this condition in the community according to age and gender, a systematic review was performed. DATA SOURCES: Data were from Embase, Medline, bibliographies of known articles and contact with experts. STUDY SELECTION: Studies were selected if data on anal incontinence could be extracted for participants over 15 years of age and living in the community. DATA EXTRACTION: Data were extracted using a piloted form on participants' characteristics, study quality and incontinence rates. DATA SYNTHESIS: Meta-analysis was used to combine data from multiple studies, and meta-regression evaluated the variation in rates according to age and gender in an analysis adjusted for study quality. RESULTS: There were 29 studies (69,152 participants), of which 5 met over half of the high quality criteria. The rate of solid and liquid faecal incontinence among people aged 15-60 years was 0.8% [95% confidence interval (CI) 0.3-1.9] in men and 1.6% (95% CI 0.8-3.1) in women. In those aged over 60, this increased to 5.1% (95% CI 3.4-7.6) in men and 6.2% (95% CI 4.9-8.0) in women. Meta-regression showed that age had a significant influence on rates of solid and liquid faecal incontinence (p = 0.007), but not gender (p = 0.368) or study quality (p = 0.085). CONCLUSIONS: The rate of solid and liquid faecal incontinence in older people is significantly higher than their younger counterparts. Gender differences in rates did not reach statistical significance.

Adolescent↗

Quadratic regression analysis for gene discovery and pattern recognition for non-cyclic short time-course microarray experiments.

BACKGROUND: Cluster analyses are used to analyze microarray time-course data for gene discovery and pattern recognition. However, in general, these methods do not take advantage of the fact that time is a continuous variable, and existing clustering methods often group biologically unrelated genes together. RESULTS: We propose a quadratic regression method for identification of differentially expressed genes and classification of genes based on their temporal expression profiles for non-cyclic short time-course microarray data. This method treats time as a continuous variable, therefore preserves actual time information. We applied this method to a microarray time-course study of gene expression at short time intervals following deafferentation of olfactory receptor neurons. Nine regression patterns have been identified and shown to fit gene expression profiles better than k-means clusters. EASE analysis identified over-represented functional groups in each regression pattern and each k-means cluster, which further demonstrated that the regression method provided more biologically meaningful classifications of gene expression profiles than the k-means clustering method. Comparison with Peddada et al.'s order-restricted inference method showed that our method provides a different perspective on the temporal gene profiles. Reliability study indicates that regression patterns have the highest reliabilities. CONCLUSION: Our results demonstrate that the proposed quadratic regression method improves gene discovery and pattern recognition for non-cyclic short time-course microarray data. With a freely accessible Excel macro, investigators can readily apply this method to their microarray data.

Algorithms↗

Regression analysis of multivariate grouped survival data.

Multivariate failure time data arise when each study subject may experience several types of event or when there are clusterings of observational units such that failure times within the same cluster are correlated. The failure times are often subject to interval grouping or have truly discrete measurements. In this paper, the marginal distribution for each discrete failure time variable is formulated by a grouped-data version of the proportional hazards model while the dependence structure is unspecified. Generalized estimating equations in the spirit of Liang and Zeger (1986, Biometrika 73, 13-22) are proposed to estimate the regression parameters and survival probabilities. The resulting estimators are consistent and asymptotically normal. Robust estimators for the limiting covariance matrices are constructed. Simulation studies demonstrate that the asymptotic approximations are adequate for practical use and that ignoring the intracluster dependence in the variance-covariance estimation would lead to invalid statistical inference. A psychological experiment is provided for illustration.

Age Factors↗

Exosurf versus Survanta surfactant preparations: proportional-hazards regression analysis of time to successful extubation and discontinuation of oxygen therapy.

Time to successful extubation (VENT-DC) and time to discontinuation of oxygen therapy (OX-DC) for infants treated with Exosurf and Survanta surfactant preparations were compared in an observational study. Proportional-hazards (Cox) regression was used to adjust for confounding by differences in baseline risk factors between treatment groups. The Exosurf/Survanta hazard ratios, representing the relative instantaneous probabilities of the next event (VENT-DC or OX-DC) occurring in the Exosurf versus Survanta groups, were close to the null value of 1.0 and were not statistically significant. For VENT-DC the ratio was 1.08 (95% confidence interval 0.73 to 1.61, p = 0.60); for OX-DC the ratio was 0.87 (0.58 to 1.30, p = 0.83). By contrast, the hazard ratios for female versus male sex and black versus white race, presented as internal reference values, were further from the null value and were statistically significant. Survival curves showing the estimated proportion of patients in whom VENT-DC or OX-DC was achieved over time were generated from the proportional hazards models. Our findings suggest that any difference between Exosurf and Survanta surfactant for VENT-DC or OX-DC is likely considerably smaller than the differences associated with gender or race. The rationales for (1) using proportional-hazards regression when duration of therapy is an outcome of interest and (2) generating internal reference risk ratios when multivariate analysis is used are discussed in relation to future studies of hyaline membrane disease or other problems of prematurity.

Aged↗

On the use of regression analysis for the estimation of human biological age.

The present investigation compares three linear regression procedures for the definition of human biological age (bioage). As a model system for bioage definition is used the variations with age of blood hemoglobin (B-hemoglobin) in males in the age range 50-95 years. The bioage measures compared are: 1: P-bioage; defined from regression of chronological age on B-hemoglobin results. 2: AC-bioage; obtained by indirect regression, using in reverse the equation describing the regression of B-hemoglobin on age in a reference population. 3: BC-bioage; defined by orthogonal regression on the reference regression line of B-hemoglobin on age. It is demonstrated that the P-bioage measure gives an overestimation of the bioage in the younger and an underestimation in the older individuals. This 'regression to the mean' is avoided using the indirect regression procedures. Here the relatively low SD of the BC-bioage measure results from the inclusion of individual chronological age in the orthogonal regression procedure. Observations on male blood donors illustrates the variation of the AC- and BC-bioage measures in the individual.

Aged↗

Lung cancer mortality among workers in the European production of man-made mineral fibers--a Poisson regression analysis.

One hundred and eighty-one lung cancer deaths among workers during 301,085 person-years in European man-made mineral fiber production between 1930-1955 and 1982 were analyzed according to Poisson regression models including age, calendar period, country, and exposure variables. Time since first employment was the variable most strongly associated with lung cancer risk in both the rock-slag wool and glass wool subcohorts. Workers in the early technological phase were at higher risk than those in other categories, particularly in rock-slag wool production. No clear trend with duration of employment was suggested. No major changes occurred in the interpretation of the results when workers with less than one year of employment or less than 20 years since first exposure were excluded. The original results, based on analyses for standardized mortality ratios, were confirmed, and workers with a short duration of employment or a short time since first employment did not need to be excluded from the analysis.

Air Pollutants, Occupational↗

Regression analysis of overdispersed correlated count data with subject specific covariates.

A robust likelihood approach for the analysis of overdispersed correlated count data that takes into account cluster varying covariates is proposed. We emphasise two characteristics of the proposed method: That the correlation structure satisfies the constraints on the second moments and that the estimation of the correlation structure guarantees consistent estimates of the regression coefficients. In addition we extend the mean specification to include within- and between-cluster effects. The method is illustrated through the analysis of data from two studies. In the first study, cross-sectional count data from a randomised controlled trial are analysed to evaluate the efficacy of a communication skills training programme. The second study involves longitudinal count data which represent counts of damaged hand joints in patients with psoriatic arthritis. Motivated by this study, we generalize our model to accommodate for a subpopulation of patients who are not susceptible to the development of damaged hand joints.

Arthritis, Psoriatic↗

Poisson regression analysis of the mortality among a cohort of World War II nuclear industry workers.

A historical cohort mortality study was conducted among 28,008 white male employees who had worked for at least 1 month in Oak Ridge, Tennessee, during World War II. The workers were employed at two plants that were producing enriched uranium and a research and development laboratory. Vital status was ascertained through 1980 for 98.1% of the cohort members and death certificates were obtained for 96.8% of the 11,671 decedents. A modified version of the traditional standardized mortality ratio (SMR) analysis was used to compare the cause-specific mortality experience of the World War II workers with the U.S. white male population. An SMR and a trend statistic were computed for each cause-of-death category for the 30-year interval from 1950 to 1980. The SMR for all causes was 1.11, and there was a significant upward trend of 0.74% per year. The excess mortality was primarily due to lung cancer and diseases of the respiratory system. Poisson regression methods were used to evaluate the influence of duration of employment, facility of employment, socioeconomic status, birth year, period of follow-up, and radiation exposure on cause-specific mortality. Maximum likelihood estimates of the parameters in a main-effects model were obtained to describe the joint effects of these six factors on cause-specific mortality of the World War II workers. We show that these multivariate regression techniques provide a useful extension of conventional SMR analysis and illustrate their effective use in a large occupational cohort study.

Adult↗

Use of regression analysis to explain the variation in prescribing rates and costs between family practitioner committees.

There are proposals to set up prescribing budgets for family practitioner committees (now family health services authorities) and indicative prescribing amounts for practices. An intelligible model is therefore required for specifying budgetary allocations. Regression analyses were used to explain the variation in prescription rates and costs between the 98 family practitioner committees of England and Wales in 1987. Fifty one per cent of the variation in prescription rates and 44% of the variation in prescription costs per patient could be explained by variations in the age-sex structure of family practitioner committees. The standardized mortality ratio for all causes and patients in 1987, and the number of general practice principals per 1000 population in 1987, but not the Jarman under-privileged area score were found to improve the predictive power of the regression models significantly (P less than 0.01). The predictions of the model for the 10 family practitioner committees with the highest and lowest prescription rates or costs are reported and discussed. Potential improvements in models of prescribing behaviour may be thwarted by two problems. First, the paucity of readily available data on health care need at family practitioner committee and practice levels, and secondly, the increasing complexity in the statistical techniques required may render the procedure less intelligible, meaningful and negotiable in a contentious field.

Costs and Cost Analysis↗