Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Statistical Regression”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 235 records · Page 13Linked to original sources
PubMed2004

How children with special health care needs affect the employment decisions of low-income parents.

OBJECTIVES: To better understand the impact of having a child with special health care needs (CSHCN), on low-income parents' employment decisions. METHODS: Using data from the 1999 and 2000 National Health Interview Survey (NHIS), we estimate multivariate statistical regressions (logit and tobit models) to estimate the relationship between having a CSHCN and the likelihood of employment and hours of employment for a sample-of low-income single parents. RESULTS: Controlling for differences in demographic and family characteristics, we find no significant association between having a CSHCN and the probability of work or the number of hours worked among low-income single-parent families. Separate analysis of different dimensions of special health care needs shows that parents of children with activity limitations are significantly less likely to work and work fewer hours. This result does not hold true for the group of children defined based on elevated or special service use, or for the group of children with specific chronic conditions. CONCLUSIONS: These results indicate that only a specific subset of children with special needs present difficulties for low-income parents' work. This suggests that policies to help low-income single parents of children with disabilities move into work should target this specific subset of children with special health care needs.

Adolescent↗
PubMed2003

Individual unionization decisions: a multilevel model of cost-benefit influences.

An important part of the unionization process is predicting the individual's vote for or against union representation. We proposed and tested a multilevel model based on the relative importance of costs and benefits of representation. Regression statistics from within-person analyses were used to show the influence of perceived costs and benefits on judgment policies about intent to vote in a representation election. Using these statistics as outcome variables, between--person analyses were used to show the influence of decision frame on cost-benefit influences, in which framing a vote as one for or against the union was conceived as a contextual variable in an election. Results were used to extend prior unionization research and to suggest how employers and unions may attend to framing effects in an election.

Adolescent↗
PubMed1999

Prevalence of overweight and obesity among school children in Jena (Germany).

OBJECTIVES: To examine the prevalence and changes in the prevalence of overweight and obesity among school children in Jena (Germany) in the last twenty years and to identify factors associated with childhood overweight. DESIGN: Cross-sectional surveys in 1975, 1985, 1995 and a household questionnaire in 1995. SUBJECTS: Children from schools in Jena, aged 7-14 y, participated (1975 : 1002 boys and 1000 girls; 1985 : 781 boys and 753 girls; 1995 : 989 boys and 912 girls). MEASUREMENTS: Prevalence of overweight or obesity based on the 90th or 97th age- and sex-specific percentile of the body mass index (BMI) developed for French children. In 1995 factors examined in relation to overweight included birth weight, birth length, age-class, number of children in household, occupation of the father, education of the mother and size of flat (apartment). RESULTS: In boys the prevalence of overweight increased from 10.0 to 16.3% and in girls from 11.7 to 20.7% between 1975 and 1995. The prevalence of obesity increased from 5.3 to 8.2% in boys and from 4.7 to 9.9% in girls between 1975 and 1995. However, the peak in the increase of overweight as well as of obesity lie for both sexes between 1985 and 1995. Using logistic regression, statistically significant associations with overweight were found for occupation of the father, birth weight in both sexes and additionally, for size of flat in girls. CONCLUSIONS: Overweight and obesity are increasing health problems among Jena children. Further investigations are needed to explore the influence of factors such as feeding pattern, food habits and physical activity on overweight. Special attention should be paid to the further social development in the society and to the link between low social class and overweight. Through such investigations effective preventive strategies could be developed.

Adolescent↗
PubMed2006

Effect of magnitude and timing of maternal pregnancy blood lead (Pb) levels on birth outcomes.

OBJECTIVE: Associations between magnitude and timing of maternal pregnancy blood lead (Pb) levels (BLLs), birth weight, and total days of gestation were examined, as well as associations with related clinical diagnoses of low birth weight (LBW), preterm, and small-for-gestational-age (SGA) birth. STUDY DESIGN: Among a sample of 262 mother-infant pairs studied retrospectively, one-way analysis of variance and regression statistics were used to measure the relationship between level of maternal pregnancy BLLs and birth outcomes while controlling for key maternal and newborn factors. RESULTS: Women with maximum pregnancy BLLs (max-PBLLs) > or =10 microg/dl tended to give birth earlier and their babies were at substantially increased risk for preterm and SGA birth. By holding other explanatory factors constant, each unit increase in max-PBLL above 10 mug/dl was found to be associated with a decrease of -0.3 in total days of gestation. Compared to women with lower levels, women with max-PBLLs > or =10 microg/dl were at a threefold increased risk for preterm birth (adjusted OR=3.2, 95% CI 1.2-7.4) and more than a fourfold increased risk for having an SGA infant (adjusted OR=4.2, 1.3-13.9). Second trimester maximum BLLs > or =10 microg/dl were associated with a steep decrease in total days of gestation (a decrease of -1.0 days per each unit increase above 10 microg/dl). CONCLUSIONS: These data provide evidence of the adverse effects of maternal pregnancy BLLs, particularly when levels are > or =10 microg/dl. Prenatal Pb exposure at these levels was associated with significant decreases in total days of gestation and an increased risk of preterm and SGA birth.

Adult↗
PubMed1995

Combination plots as graphical tools in the study of enzyme inhibition.

Although statistical regression has become the method of choice in the analysis of enzyme kinetics, graphical methods continue to be useful on account of their illustrative capabilities. It is pointed out in this paper that enzyme inhibition data may be presented more efficiently as a single linear plot than the traditional way as a family of lines. This approach has been taken previously by Hunter and Downs [Hunter and Downs (1945) J. Biol. Chem. 157, 427-446] but has remained neglected. A new version of this type of plot (combination plot) has been devised which is linear for competitive, non-competitive, uncompetitive and linear mixed inhibition and has a characteristic appearance for each type of inhibition behaviour. The slopes and intercepts not only indicate directly the dissociation constant but also provide quantitative criteria for the nature of inhibition. This plot should serve as a useful graphical tool in enzyme research as well as in biochemical education.

Binding, Competitive↗
PubMed2003

Greater reductions in A1C in type 2 diabetic patients new to therapy with glyburide/metformin tablets as compared to glyburide co-administered with metformin.

BACKGROUND: A cohort of patients with type 2 diabetes, prescribed glyburide/metformin tablets, experienced significantly greater improvements in glycaemic control compared to patients receiving glyburide co-administered with metformin. AIM: To compare the change in A1C for type 2 diabetic patients new to combination therapy with fixed-dose glyburide/metformin tablets vs. glyburide co-administered with metformin in a usual-care setting. METHODS: This retrospective cohort study analysed medication usage via an administrative pharmacy claims database and the patients' corresponding laboratory results. Patients were new to antidiabetic combination therapy with glyburide/metformin tablets or glyburide co-administered with metformin between August 2000 and July 2001 and had A1C measurements at baseline and within 76-194 days of initiating combination therapy. The change from baseline in A1C was analysed using statistical regression to adjust for significant covariates (baseline A1C and dosage). Adherence with therapy was also compared. RESULTS: The cohort consisted of 950 patients who received glyburide/metformin tablets and 471 taking glyburide co-administered with metformin. Glyburide/metformin patients were younger (mean age = 56 vs. 60 years, p < 0.0001) and received lower doses of each drug than patients taking glyburide co-administered with metformin (glyburide mean final dose = 6 vs. 10 mg/day, p < 0.0001; metformin = 893 vs. 1297 mg/day, p < 0.0001). The mean decrease from baseline A1C, adjusted for baseline A1C and dosage, of 2.02% for glyburide/metformin tablets was significantly (p < 0.0001) greater than the decrease of 1.49% for glyburide co-administered with metformin. Glyburide/metformin patients with baseline A1C >/= 8 experienced a significantly (p < 0.0001) greater decrease in A1C of 2.93% compared to 1.92% for glyburide co-administered with metformin. For patients with baseline A1C < 8%, the difference between the A1C responses remained significant, even though the reductions in A1C were smaller for both glyburide/metformin tablets and glyburide co-administered with metformin (0.54% and 0.23%, p = 0.0017). Patients were more adherent with glyburide/metformin tablets (84% vs. 76%, p < 0.0001), though regression analysis indicated that adherence was not a significant predictor of change in A1C. CONCLUSIONS: The lower medication doses delivered by glyburide/metformin tablets provided a significantly greater reduction in A1C than did glyburide co-administered with metformin in patients with type 2 diabetes, especially when baseline A1C >/= 8%.

Adult↗
PubMed2003

Predicting GFR in children and adults: a comparison of the Cockcroft-Gault, Schwartz, and modification of diet in renal disease formulas.

BACKGROUND: A rapid prediction of glomerular filtration rate (GFR) is often needed in clinics. Formulas based on plasma creatinine level are being increasingly used, Schwartz for children, supposed to give GFR; Cockcroft-Gault for adults, supposed to indicate the creatinine clearance; and a recent formula introduced by the Modification of Diet in Renal Disease (MDRD) group. Our objective was to test whether one single formula could suffice and which one gives the best estimation of GFR. METHODS: In 198 children (with two kidneys, single kidney, or renal transplant) and 116 adults (single kidney and transplanted), we measured inulin clearance and creatinine clearance and calculated Cockcroft-Gault, MDRD and, in children only, Schwartz. Data were compared with analysis of variance (ANOVA), regression statistics, and concordance studies. RESULTS: In patients over 12 years of age, Cockcroft-Gault was almost similar to GFR corrected for body surface and creatinine clearance exceeded GFR by more than 20%; Schwartz was above creatinine clearance excepted for transplanted children. In younger children, no prediction approached GFR. Predictions were well correlated with GFR, but concordance studies showed Schwartz with dispersed results and GFR overestimated (20 mL/min/1.73 m2); Cockcroft-Gault was close to GFR and results were dispersed; MDRD in children gave a large overestimation and badly dispersed results; in transplanted adults its prediction was good. CONCLUSION: Cockcroft-Gault prediction could be used for children over 12 years of age and adults; it should not be considered as creatinine clearance but as GFR corrected for body surface, it is merely a prediction, 95% of the results are between +/-40 mL/min/1.73 m(2) in children and +/-30 mL/min/1.72 m(2) in adults. In younger children no formula is satisfying.

Adolescent↗
PubMed2002

One-year mortality among unselected outpatients with heart failure.

OBJECTIVE: To estimate 1-year mortality and prognostic factors in unselected outpatients with heart failure, and to compare the observed mortality with the estimates of the primary care physicians. METHODS AND RESULTS: Four hundred and eleven consecutive patients with heart failure New York Heart Association (NYHA) class II-IV (mean population age 75 years, 56% males) were enrolled in 71 primary care offices throughout Switzerland. During a mean follow-up period of 1.4 years, 68 patients had died. One-year total mortality was 12.6% compared to 4.3% in the underlying Swiss population (standardized mortality ratio 3.0). Among patients with heart failure NYHA II, III and IV, mortality was 7.1%, 15.0% and 28.0%, respectively. In multivariate Cox regression, statistically significant (P<0.05) predictors of mortality were NYHA class (NYHA III: risk ratio [RR]=1.6; NYHA IV: RR=2.2), recent hospital stay for heart disease (RR=2.3), creatinine>120 micromol.l(-1) (RR=1.8) systolic blood pressure<100 mmHg (RR=2.4), heart rate>100 min(-1) (RR=2.7), age (per 10 years, RR=1.6) and female gender (RR=0.49). Among patients with reduced left ventricular ejection fraction, 1-year mortality was 14.3%, and predictors were similar except that female gender was no longer associated with reduced mortality. Primary care physicians significantly overestimated 1-year mortality (estimated mortality 25.9% vs observed mortality 12.6%,P =0.001). CONCLUSIONS: Unselected outpatients with heart failure have a poor prognosis, particularly those with advanced heart failure and a recent hospital stay for heart disease. Primary care physicians are aware of the high mortality of this growing patient population.

Aged↗
PubMedDate not supplied

[Creatine-kinase MM in the perinatal period].

One hundred eight newborns (gestational age 36 + 1.8 weeks and birth weight 2860 +/- 240 g) had muscle type Creatine-Kinase activity (CK-MM) assayed immediately after birth (CK-I) and serially at 6-10 h (CK-II), 20-30 h (CK-III) and 40-60 h (CK-IV) of age. Using statistical Regression analysis, CK-MM levels were correlated to four perinatal parameters: gestational age, birth weight, neonatal acidosis (pH less than 7.15), mode of delivery (vaginal/cesarean section). It was observed that CK-MM activity was dependent on gestational age and correlated to the mode of delivery at time III (600 U/l) and IV (400 U/l) comparing the mean serum values of 156 + 44 U/l observed in atraumatic delivery (p less than or equal to 0.05). Our results demonstrate markedly higher levels of CK-MM following vaginal delivery especially if complicated by forceps, vacuum and breech presentation, suggesting that birth trauma may be responsible for this phenomenon.

Acidosis↗
PubMed2005

Bayesian coclustering of Anopheles gene expression time series: study of immune defense response to multiple experimental challenges.

We present a method for Bayesian model-based hierarchical coclustering of gene expression data and use it to study the temporal transcription responses of an Anopheles gambiae cell line upon challenge with multiple microbial elicitors. The method fits statistical regression models to the gene expression time series for each experiment and performs coclustering on the genes by optimizing a joint probability model, characterizing gene coregulation between multiple experiments. We compute the model using a two-stage Expectation-Maximization-type algorithm, first fixing the cross-experiment covariance structure and using efficient Bayesian hierarchical clustering to obtain a locally optimal clustering of the gene expression profiles and then, conditional on that clustering, carrying out Bayesian inference on the cross-experiment covariance using Markov chain Monte Carlo simulation to obtain an expectation. For the problem of model choice, we use a cross-validatory approach to decide between individual experiment modeling and varying levels of coclustering. Our method successfully generates tightly coregulated clusters of genes that are implicated in related processes and therefore can be used for analysis of global transcript responses to various stimuli and prediction of gene functions.

Algorithms↗
PubMed2006

Predictors of anxiety and depression following pregnancy termination: a longitudinal five-year follow-up study.

BACKGROUND: The aims of the study were to assess anxiety and depression in women who had experienced either a miscarriage or an induced abortion, to compare the women's level of distress with that of a general population sample, and to find predictors of anxiety and depression six months and five years after the event. METHODS: A prospective, longitudinal follow-up study. Women who experienced miscarriage (n = 40) and induced abortion (n = 80) were interviewed ten days (T1), six months (T2), two years (T3), and five years (T4) after the event. On each occasion, they completed the Hospital Anxiety and Depression Scale and the Life Events Scale. Paired-sample t-test, logistic regression, and multiple linear regression statistical tests were used. RESULTS: Women with miscarriage had significantly more anxiety and depression at T1 than the general population, while women with induced abortion had significantly more anxiety at all time points and more depression at T1 and T2. In both groups, important predictors of anxiety and depression at T2 and T4 were recent life events and poor former psychiatric health. Childbirth events between T1 and T4 had no significant influence on the scores. For women with induced abortion, doubt about the decision to abort was related to depression at T2 (p <0.05), while a negative attitude towards induced abortion was associated with anxiety at T2 (p <0.05) and T4 (p <0.05). CONCLUSION: Correlates of anxiety and depression may be used to better identify women who are at risk of negative psychological responses following pregnancy termination.

Abortion, Induced↗
PubMed1996

Predicting endurance limits in arm cranking exercise with a subjectively based method.

The purpose of this study was to compare a method for predicting endurance capacity at different workloads based on subjective perception of effort with the technique for determining 'critical power' (CP) proposed by Moritani et al. (1981). Three trained paraplegic subjects performed two protocols during separate sessions. The 'subjective' protocol consisted of graded and continuous tests on an arm crank ergometer, during which cardiorespiratory (heart rate, oxygen consumption, ventilation and respiratory exchange ratio) and subjective (Borg's 10-point scale) parameters were monitored continuously. A statistical regression analysis for power functions was performed to obtain the individual ¿iso-perception' curves in a power/duration reference system. The other protocol consisted of the determination of the CP of the muscle groups involved in arm cranking exercises according to the protocol proposed by Moritani et al. (1981). Subjects performed three tests in which the power output remained constant (50, 37.5 and 25 W) and led to the onset of muscular fatigue. The results of the two protocols were compared. The validity of the subjective perception of effort as a predictor of individual endurance capacity was confirmed by the relationship between time to exhaustion and work performed during constant-power tests as proposed by Moritani et al. (1981).

Adult↗
PubMed1993

Accelerated forgetting in Alzheimer-type dementia.

Accelerated forgetting of name-face associations and grocery list items within the first hour postpresentation is demonstrated in 80 persons with Alzheimer's disease (AD) compared to 80 control subjects matched on age, education, and gender. Differences in forgetting which exceeded statistical regression effects remained, even when AD and control subjects were matched on rate of acquisition during the learning trials of name-face associations. Results are discussed in relation to the neuropathology of AD, organic amnestic disorders, and methodological factors concerning previous research on forgetting in persons with AD.

Aged↗
PubMed1999

Evaluation of a new troponin I method on the Bayer Immuno 1 immunoassay analyser.

We have evaluated the analytical and clinical performance of an automated immunoassay for serum cardiac troponin I (Bayer Immuno 1TM, Bayer Diagnostics, Tarrytown, NY). The between batch imprecision was found to be between 1.2 and 3.2% over the concentration range 2.5 - 34.0 microg/L. The analytical range obtained from duplicate analysis of patient samples and defined as a coefficient of variation of 10% or less was 0.3 - 200 microg/L. The detection limit was found to be less than 0.1 microg/L. A method comparison with the Dade Stratus method (Dade Behring, Wilmington, DE) yielded regression statistics with a slope of 0.705 and an intercept of -0.260. An analysis of samples from 40 patients with renal failure demonstrated six with detectable levels of troponin I (0.2 - 1.9 microg/L). Samples from patients with paraproteinaemia did not demonstrate detectable troponin I (from n = 30); however, two patients with elevated rheumatoid factor titers (from n = 20) demonstrated a detectable amount of troponin I (0.1 and 0.2 microg/L). In a study of 100 patients admitted with acute chest pain and a diagnosis of unstable angina, 6 were subsequently diagnosed as having suffered a myocardial infarction. On admission the sensitivity and specificity of the troponin I results were 26.7% and 94.7%, respectively, moving to 100% and 83% 12 hours after admission.

Analysis of Variance↗
PubMed1993

Students' knowledge and behavior concerning safer sex: a UK study.

Most students in institutions of further and higher education in the northeast of England who responded to a 1989 survey dealing with sexual behavior and safer sex were heterosexual and had been sexually active or intended to have sexual intercourse. Many students, however, believed that safer sex implied having no sex at all or were unaware of the role of nonpenetrative sex in preventing human immunodeficiency virus (HIV) infection. They did not use condoms, in spite of having positive attitudes about condom use, and they engaged in one or more sexual activities that increased the risk of acquiring HIV infection. In view of the increasing incidence of acquired immunodeficiency syndrome (AIDS), the authors assert, these findings are cause for concern and require concerted action by health educators and society to encourage safer sex practices and to prevent the spread of HIV infection.

Acquired Immunodeficiency Syndrome↗
PubMed1997

The Australian study of HIV and injecting Drug Use. Part II: predicting exposure to hepatitis C and hepatitis B.

Researchers agree that while hepatitis B maybe in control, hepatitis C is present in epidemic proportions among injecting drug users and that current HIV prevention strategies have not been sufficient to halt the spread of this hepatitis virus, although there is some evidence to suggest that incidence rates are stabilizing. Since there is no effective cure and it is unlikely that a vaccine will become available in the foreseeable future all efforts to control the spread of hepatitis C must rely on education and prevention strategies. The Australian Study of HIV and Injecting Drug Use is a cross-sectional national study designed to investigate exposure to and risks for infection with bloodborne viruses. Of those volunteering a usable blood sample for hepatitis C antibody and hepatitis B core antibody testing 55% and 19%, respectively, returned reactive test results. Logistic regression statistical models were used to identify risk factors for hepatitis C and hepatitis B. Risk factors for hepatitis C were identified as duration of use, use of opiates on last injecting occasion, education level, treatment status and having a history of sexually transmissible diseases. Risk factors associated with hepatitis B were duration of use, and use of opiates on last injecting occasion. The lack of identifiable risk factors for hepatitis B suggest that past rather than current injecting and sexual behaviour patterns are required to predict accurately risk of exposure to hepatitis B. In addition to this, one-third of respondents reported being vaccinated against hepatitis B. Respondents perceived themselves to be at greater risk from hepatitis C than from hepatitis B or HIV. A discussion of strategies needed to prevent the spread of the hepatitis viruses will be presented along with recommendations for further research.

Journal Article↗
PubMed2004

Time variant exposure analysis (TVEA): a measurement tool for characterizing particulate exposure determinants in construction.

A work sampling-based approach, time variant exposure analysis (TVEA), was developed for assessment of determinants for particulate air contaminants in dynamic construction environments. To use TVEA, the field researcher records observations at fixed intervals to systematically survey over 30 potential determinants that could affect exposure to three types of particulate matter: quartz-containing dusts, diesel exhaust, and a general grouping of "other particles" that includes welding fume and wood dust. Two field studies were conducted to address questions of inter-rater reliability (n = 20) and coding interval appropriateness (n = 21) for the TVEA method. At least substantial inter-rater agreement (kappa > 0.60) was obtained for the TVEA variables related to tool or machine use, process, material, source intensity, and source orientation. Kappa values for source direction (0.22-0.38) and number of sources (0.38-0.60) showed comparatively lower agreement for all particulate types. Observation interval appropriateness was analyzed using linear regression to compare a 5-min observation interval "gold standard" with alternate intervals. Regression statistics indicated that while 30 min is an acceptable interval for exposure assessment, 15 min optimizes precision and practicality by ensuring that 95% of all observations differ less than ten percentage points from the "true" values. TVEA is a useful exposure assessment tool for the dynamic construction environment. It is flexible in that only those determinants that are of interest need be coded and the coding interval can be adjusted to accommodate the level of precision desired.

Air Movements↗