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.

SourceResearch connectionFrom yearThrough year

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

At least 1,279 records · Page 71Linked to original sources

PubMed2003

Effects of exposure measurement error when an exposure variable is constrained by a lower limit.

Epidemiologic studies routinely suffer from bias due to exposure measurement error. In this paper, the authors examine the effect of measurement error when the exposure variable of interest is constrained by a lower limit. This is an important consideration, since often in epidemiologic studies an exposure variable is constrained by a lower limit such as zero or a nonzero detection limit. In this paper, attenuation of exposure-disease associations is defined within the framework of a classical model of uncorrelated additive error. Then, the special case of nonlinearity due to the effect of a lower threshold is examined. A general model is developed to characterize the effect of random measurement error when there is a lower threshold for recorded values. Findings are illustrated under the assumption that the true exposure follows the lognormal and gamma distributions. The authors show that the direction and magnitude of bias in estimated exposure-response associations depends on the population distribution of the exposure, the magnitude of the recording threshold, the value assigned to below-threshold measurement results, and the variance in the measured exposure due to random measurement error.

Bias↗

PubMed2005

An approach to estimate between- and within-group correlation coefficients in multicenter studies: plasma carotenoids as biomarkers of intake of fruits and vegetables.

In a multicenter study, the overall correlation between two variables can be broken down into a within- and a between-group correlation reflecting associations at the individual and aggregate levels, respectively. A random-effects model is used to estimate variance components of nutrition-related variables and the within- and between-group correlation coefficients. Using data from the European Prospective Investigation into Cancer and Nutrition (EPIC), the authors analyzed the association between levels of three plasma carotenoids (alpha-carotene, beta-cryptoxanthin, and lycopene) and dietary intake of three items (total fruits, carrots, and tomatoes), assessed through dietary questionnaire and single 24-hour dietary recall measurements, in a cross-sectional study involving 3,089 subjects from nine European countries. Intraclass correlation coefficients were 0.178 for alpha-carotene, 0.216 for beta-cryptoxanthin, and 0.299 for lycopene. The between-group correlation coefficients were higher than the within-group coefficients for all three carotenoids. For beta-cryptoxanthin and fruit intake, the between-group versus the within-group correlations were 0.78 and 0.26 for the dietary questionnaire and 0.85 and 0.19 for the 24-hour dietary recall. Results indicate that variability of exposure is driven mainly by the individual compared with the aggregate variation and that biomarker levels perform fairly accurately in discriminating population-level consumption of fruits and vegetables.

Adult↗

PubMed1991

Pulse rate and sodium intake interact to determine blood pressure. A population study.

It has been suggested that certain factors need to be present for the pressor effect of sodium to become apparent. The present population study investigated whether pulse rate (PR) and salt intake interact to determine blood pressure (BP). Twenty-four hour urinary sodium (UNaV) was used as a measure of salt intake. A random population sample, including 2081 subjects with a minimum age of 18 years, was stratified on tertiles of PR. In subjects with a slow PR(less than 68 beats/min) and in those with a fast PR (greater than 78 beats/min), a significant curvilinear relationship between BP and UNaV was found, while in subjects with an intermediate PR the BP-UNaV correlation was not significant. There was also a significant interaction between PR and UNaV, indicating that when UNaV increased from 160 to 300 mmol/24 h, systolic/diastolic BP rose by 2/1 mm Hg in the fast PR third, but declined by 1/1 mm Hg in the slow PR third. These divergent trends could not be explained by smoking habits or alcohol consumption and were still present after cumulative adjustment for other important correlates of BP, ie, gender, age, body weight, urinary potassium, and contraceptive pill intake in women. In conclusion, the present data suggest that in subjects with a pulse rate greater than 78 beats/min a high salt intake may be associated with blood pressure elevation, whereas the opposite tendency is observed in individuals with a pulse rate less than 68 beats/min.

Adolescent↗

PubMed2004

A perturbation-based method for calculating explicit likelihood of evolutionary co-variance in multiple sequence alignments.

MOTIVATION: The constituent amino acids of a protein work together to define its structure and to facilitate its function. Their interdependence should be apparent in the evolutionary record of each protein family: positions in the sequence of a protein family that are intimately associated in space or in function should co-vary in evolution. A recent approach by Ranganathan and colleagues proposes to look at subsets of a protein family, selected for their sequence at one position, to see how this affects variation at other positions. RESULTS: We present a quantitative algorithm for assessing covariation with this approach, based on explicit likelihood calculations. By applying our algorithm to 138 Pfam families with at least one member of known structure, we demonstrate that our method has improved power in finding physically close residues in crystal structures, compared to that of Ranganathan and colleagues. SUPPLEMENTARY INFORMATION: www.afodor.net/bioinfosup.html

Algorithms↗

PubMed2006

Transcriptome network component analysis with limited microarray data.

UNLABELLED: Network component analysis (NCA) is a method to deduce transcription factor (TF) activities and TF-gene regulation control strengths from gene expression data and a TF-gene binding connectivity network. Previously, this method could analyze a maximum number of regulators equal to the total sample size because of the identifiability limit in data decomposition. As such, the total number of source signal components was limited to the total number of experiments rather than the total number of biological regulators. However, networks that have less transcriptome data points than the number of regulators are of interest. Thus it is imperative to develop a theoretical basis that allows realistic source signal extraction based on relatively few data points. On the other hand, such methods would inherently increase numerical challenges leading to multiple solutions. Therefore, solutions to both the problems are needed. RESULTS: We have improved NCA for transcription factor activity (TFA) estimation, based on the observation that most genes are regulated by only a few TFs. This observation leads to the derivation of a new identifiability criterion which is tested during numerical iteration that allows us to decompose data when the number of TFs is greater than the number of experiments. To show that our method works with real microarray data and has biological utility, we analyze Saccharomyces cerevisiae cell cycle microarray data (73 experiments) using a TF-gene connectivity network (96 TFs) derived from ChIP-chip binding data. We compare the results of NCA analysis with the results obtained from ChIP-chip regression methods, and we show that NCA and regression produce TFAs that are qualitatively similar, but the NCA TFAs outperform regression in statistical tests. We also show that NCA can extract subtle TFA signals that correlate with known cell cycle TF function and cell cycle phase. Overall we determined that 31 TFs have statistically periodic TFAs in one or more experiments, 75% of which are known cell cycle regulators. In addition, we find that the 12 TFAs that are periodic in two or more experiments correspond to well-known cell cycle regulators. We also investigated TFA sensitivity to the choice of connectivity network we constructed two networks using different ChIP-chip p-value cut-offs. AVAILABILITY: The NCA Toolbox for MATLAB is available at http://www.seas.ucla.edu/~liaoj/download.htm.

Algorithms↗

PubMed2000

Comorbidity of five chronic health conditions in elderly community residents: determinants and impact on mortality.

BACKGROUND: Comorbidity is common in elderly persons. Its extent, correlates, and life-threatening impact in representative community residents are unclear. METHODS: Self-reported information of physician-diagnosed coronary artery disease (CAD), cerebrovascular disease (CVD), diabetes, and cancer was obtained annually between 1986-87 and 1992-93, and hypertension was obtained triennially from the participants of the Duke Established Populations for Epidemiologic Studies of the Elderly, a stratified multistage sample of 4,126 Black and White community residents aged 65-100, living in a five-county area of North Carolina. Date of death was obtained from death certificates identified through search of the National Death Index. Statistical procedures included descriptive statistics, logistic regression, and survival analysis. RESULTS: Of this sample, 57% reported hypertension, 20% diabetes, 15% CAD, 9% cancer, and 9% CVD; 29% reported none of these conditions, whereas 29% reported two or more. Demographic characteristics were not related to comorbidity with CVD or cancer. Increased education tended to be protective. The effect of age, gender, and race varied with condition. At baseline there was substantial comorbidity among hypertension, CAD, CVD, and diabetes, but not with cancer. Hypertension, CVD, and diabetes were risk factors for CAD, whereas diabetes was a risk factor for CVD. After controlling for demographic characteristics, all health conditions except hypertension were predictive of 6-year mortality, as was the presence of comorbidity. CONCLUSION: We found significant comorbidity in older persons who have hypertension, CAD, CVD, or diabetes; particular risk of developing comorbidity, particularly CAD, among those with hypertension, CVD, and diabetes; and risk of CVD in those with diabetes. With the exception of hypertension, these conditions, and comorbidity per se, are life-threatening.

Age Factors↗

PubMed1988

Who consults and where? Sociocultural differentiation in access to health care in urban Africa.

Sociocultural differentiation in health behaviour was studied among 500 mothers randomly chosen in the suburban area of Dakar, Senegal. Variables were age, marital status, educational level, socioeconomic conditions, urban experience, town integration, village attachment, social insurance, ethnic group and disease categories. Access to health care was considered for the last illness of the mother and her youngest child under five. Town integration and disease category (and social insurance for children) were the only variables correlated with the number of consultations. Socioeconomic level and social insurance (and educational level for mothers) were associated with preference for private rather than public health systems.

Adult↗

PubMed1990

The comparison of trends in perinatal mortality in small areas.

Year to year changes in perinatal mortality rates for small populations are difficult to interpret. Because local rates are based on small numbers of events, they are subject to apparently large fluctuations. A technique for analysing trends in mortality, for detecting changes in trends and comparing variations in trends between areas is described. It is illustrated by an analysis of perinatal mortality rates for Area Health Authorities in the West Midlands Region of England from 1974-1981, but the same techniques could be used to analyse trends at other times and in other places.

Birth Rate↗

PubMed1991

Demographic, social and clinical correlates of Raynaud phenomenon.

Our study has identified the demographic, social and clinical correlates of primary Raynaud phenomenon (RP) from a case-control study involving 235 subjects who were identified from a population-based prevalence survey of RP. All potential cases and a sampling of probable non-cases, identified by a screening questionnaire, were invited to a medical clinic for diagnostic testing and assignment of case or control status. Odds ratios (OR) for risk indicators for primary RP were estimated using multiple logistic regression to obtain 95% confidence intervals. The regression yielded statistically significant (p less than 0.05) positive associations for sex (OR = 3.0 for females versus males), self-reported alcohol use (OR = 1.1) and diastolic blood pressure (OR = 1.2) in those on antihypertensive medication. A significant (p = less than 0.01, OR = 0.4) negative association was found for Quetelet index (a measure of obesity). Near-significant or suggestive but non-significant negative associations were found between RP and marital status (p = 0.05, OR = 2.5), increased years of education and, for those on antihypertensive medication, higher systolic blood pressure (both p = 0.07).

Alcohol Drinking↗

PubMed1991

Oral contraceptives and breast cancer: a French case-control study.

The relationship between the risk of breast cancer and oral contraceptive use was investigated in a case-control study conducted in France between 1983 and 1987 in five public hospitals. Some 464 cases aged 25 to 56 years and 542 matched controls were interviewed about their history of the use of oral contraceptives (OC). Results are given for the entire population and for the subgroup of 358 and 379 premenopausal cases and controls. The multivariate relative risk estimate, for ever user, was 1.5 (p less than 0.01) in the whole group as well as in the premenopausal subgroup (p less than 0.02). However, there was no evidence that the effect varied appreciably according to duration of use, age at first use, use before first full-term pregnancy (FFTP) and time since first or last use. The risk was not altered for any particular brand of OC. We conclude that, because of the widespread attention given to the relationship between OC use and breast cancer, information bias might be responsible for part of the excess in risk observed among OC ever users.

Adenocarcinoma↗

PubMed1992

Seasonality in tropical AIDS: a geographical analysis.

This paper presents evidence that the growth rate of the AIDS epidemic at the district level in Uganda, Central Africa, displays a seasonally recurring geographical pattern, with epidemic acceleration in some areas of the country in the first 8 months of each year. The spatial and temporal variations in acceleration appear to be correlated with the predominant agricultural systems in different parts of Uganda. Based upon the frequently hypothesized relationship between malnourishment and the progression to clinical AIDS in HIV-infected people, it is suggested that the variations in epidemic speed reflect the seasonal patterns of nutritional deficiency which occur under some tropical agricultural systems. These preliminary findings require further verification since they have important implications for directing nutrition-related remedial responses to the AIDS epidemic in tropical countries where malnutrition and endemic HIV infection coincide.

Acquired Immunodeficiency Syndrome↗

PubMed1992

Nonoccupational determinants of blood lead concentrations in a general population.

We investigated the influence of various nonoccupational factors on blood lead levels (PbB) in a sample from the general population of southern Germany. Some 1703 men and 1661 women, aged 28-67 years, were examined in the first follow-up examination of the MONICA Augsburg cohort study in 1987-1988. Their mean PbB was 90 micrograms/l (SD:35.9) for men and 65 micrograms/l (26.4) for women. Only 5% of the men and 1% of all women exceeded a PbB level of 150 micrograms/l indicating low-level lead exposure in this population. Blood lead was significantly associated with haematocrit values (P < 0.001) and the shape of this association was curvilinear. Per gram of alcohol consumed, intake of beer had a lower impact on PbB than wine, presumably due to differential lead content in these alcoholic beverages. The alcohol-PbB associations were stronger for women than for men. The impact of smoking was generally moderate but again more prominent in women. In particular, the covariate adjusted odds ratios for women of childbearing age (28-47 years) to have PbB levels above 100 micrograms/l were 2.5 (95% confidence interval (CI): 1.3-4.7) for smoking versus non-smoking females, 2.6 (95% CI: 1.1-6.0) for women drinking up to 40 g alcohol/day compared to abstainers, and 8.9 (95% CI: 3.2-25.1) for those drinking more than 40 g alcohol/day. Other factors like age, body mass, rural place of residence, and education or job position, had only minor influences on PbB. We conclude that haematocrit values should always be considered as potential confounders in low-level lead exposure research. High alcohol consumption and cigarette smoking are strongly related to elevated blood lead concentrations in the general population and may thereby convey additional health hazards such as impaired child development or blood pressure elevations. This deserves proper public health recognition [corrected].

Adult↗

PubMed1994

Intrauterine device use and risk of tubal pregnancy: an Indonesian case-control study.

BACKGROUND: We assessed the risk of tubal pregnancy among women who (1) were currently using an intrauterine device (IUD) and (2) had discontinued IUD use while still sexually active and at risk of pregnancy using data from a multicentre case-control study of married women conducted in Indonesia. METHODS: Cases were 560 women diagnosed with histologically confirmed ectopic pregnancy from April 1989 to August 1990 at any one of 11 participating hospitals. Controls were 1120 non-pregnant women similar in age and place of residence to the cases. In-person interviews were conducted to collect information regarding current and past contraceptive use as well as other demographic and personal characteristics. RESULTS: Women currently using an IUD were considerably less likely than women not currently using contraception, but more likely than users of hormonal or surgical means of contraception, to develop a tubal pregnancy. Women who had discontinued using an IUD had a 70% subsequent increase in risk of tubal pregnancy (adjusted RR = 1.7, 95% Cl: 1.1-2.5) relative to women who had never used an IUD. This increase in risk was most pronounced in women who reported multiple episodes of IUD use and, to a lesser extent, in women with a long (> 3 year) duration of IUD use. CONCLUSIONS: The associations we observed are similar to those previously reported in studies conducted in developed countries. The results are of particular interest because this study was conducted in a location in which the Dalkon Shield IUD was never available, and among a population of married, gravid women for whom IUD use is generally considered most appropriate.

Adolescent↗

PubMed2003

A multicentre comparison of a novel surrogate marker for determining the specific potency of anti-tuberculosis drugs.

A model for evaluating the potency of a new anti-tuberculosis drug or a drug combination, based on a decline in the number of viable tubercle bacilli in patient's sputum during 5 days mono-therapy has been reported. One popular measure is based on the analysis of the decline in bacterial counts during the first 48 h of therapy and has been called early bactericidal activity (EBA). Such analyses could detect EBA for only a few drugs and were subject to variations in results obtained in different sites. To address these problems we applied a reiterative exponential decay model to evaluate the data on bacterial counts during 5 days of mono-therapy. The validity of this approach was tested using data from three previously published studies. For patients treated with isoniazid 300 mg daily, the values for the time taken to reduce the viable count by 50% (vt50) measured in days were, from a Kenyan study 0.58 days S.E.M. 0.18, from a Tanzanian study 0.41 days S.E.M. 0.04, and from a United States study 0.55 days s.e.m. 0.12. These differences were not statistically significant (P = 0.77 Kruskal-Wallis non-parametric ANOVA). Mean values of vt50 for all of the major anti-tuberculosis agents showed that there was an overlapping spectrum of activity from isoniazid 300 mg (vt50 0.58 days) to para-amino-salicylic acid (vt50 2.9 days) The variation between column means was greater than could be expected by chance (P = 0.0002 Kruskal-Wallis non-parametric ANOVA). From this, we conclude that the reiterative exponential decay model permits comparison between the data obtained in different centres and would allow the activity of a new drug to be compared with that of the currently available agents.

Algorithms↗

PubMed1986

Reductive effect of aspirin treatment on primary tumor growth and metastasis of implanted fibrosarcoma in rats.

For suppression of primary tumor growth and metastatic spread, aspirin and theophylline, either alone or combined, were given daily to inbred female BN rats after sc implantation of a syngeneic nonimmunogenic tumor. Treatment with 200 mg aspirin/kg (body wt) resulted in a statistically significant regression of tumor growth as well as of the number of metastases in the lungs. Aspirin given in a lower dose (20 mg/kg) did not show significant difference from the vehicle group. Theophylline (75 mg/kg) significantly increased primary tumor growth as well as lung metastases. Inhibition of in vitro platelet aggregation, determined in whole blood taken from non-tumor-bearing animals treated with the same therapeutic regimen, was most pronounced in those groups in which tumor growth and spread were significantly retarded. However, this positive correlation between inhibition of tumor spread and platelet aggregation was not associated with a favorable balance of prostacyclin and thromboxane A2 in these animals.

Animals↗

PubMed1989

Invasive cervical cancer and smoking in Latin America.

A case-control study of 667 patients with invasive squamous cell carcinoma of the cervix and 1,430 controls from four Latin American countries showed an age-adjusted relative risk (RR) of 1.2 [95% confidence interval (CI) = 1.0-1.4] for women who had ever smoked, with risk rising to 1.7 (95% CI, 0.8-3.6) for women who smoked greater than or equal to 30 cigarettes per day. The associations were practically eliminated after adjustment for the number of sexual partners and alcohol consumption, probably a surrogate for an unidentified life-style risk factor. Some excess risk persisted among women who smoked for extended periods (RR = 1.5 for greater than or equal to 40 yr), as well as those who began smoking at older ages (RR = 1.7 for greater than 30 yr), which suggests a late-stage effect. In addition, among women who tested positive for human papillomavirus (HPV) type 16 or 18 by filter in situ hybridization, there was an increased risk for women who had ever smoked and a dose-response relationship with the number of cigarettes smoked (adjusted RRs compared with HPV-negative nonsmokers = 5.0 for HPV-positive nonsmokers, 5.5 for less than 10 cigarettes/day, and 8.4 for greater than or equal to 10 cigarettes/day). In contrast, HPV-negative women had no increased risk associated with smoking. These results, from a high-incidence area where intensive smoking among women is still relatively rare, suggest that smoking has a limited effect on cervical cancer risk, possibly only among women with specific types of HPV.

Age Factors↗

PubMed1999

Randomized trial of fenretinide to prevent second breast malignancy in women with early breast cancer.

BACKGROUND: Fenretinide, a vitamin A analogue, has been shown to inhibit breast carcinogenesis in preclinical studies. We determined the efficacy of fenretinide in preventing a second breast malignancy in women with breast cancer. METHODS: We randomly assigned 2972 women, aged 30-70 years, with surgically removed stage I breast cancer or ductal carcinoma in situ to receive for 5 years either fenretinide orally (200 mg/day) or no treatment. The primary end point was the incidence of contralateral breast cancer or ipsilateral breast cancer 7 years after randomization. Other end points considered post hoc were the same outcomes stratified by menopausal status, incidence of distant metastases, overall mortality, and tumors in other organs. The hazards of breast cancer occurrence were determined by Cox proportional hazards regression analysis. Statistical tests were two-sided. RESULTS: At a median observation time of 97 months, there were no statistically significant differences in the occurrence of contralateral breast cancer (P =.642) or ipsilateral breast cancer (P =.177) between the two arms. However, an interaction was detected between fenretinide treatment and menopausal status in both outcomes (P for interaction in both outcomes =.045), with a possible beneficial effect in premenopausal women (contralateral breast cancer: adjusted hazard ratio [HR] = 0.66, and 95% confidence interval [CI] = 0.41-1.07; ipsilateral breast cancer: adjusted HR = 0.65, and 95% CI = 0.46-0. 92) and an opposite effect in postmenopausal women (contralateral breast cancer: adjusted HR = 1.32, and 95% CI = 0.82-2.15; ipsilateral breast cancer: adjusted HR = 1.19, and 95% CI = 0.75-1. 89). There were no statistically significant differences between the two arms in tumors in other organs, incidence of distant metastasis, and all-cause mortality. CONCLUSIONS: Fenretinide treatment of women with breast cancer for 5 years appears to have no statistically significant effect on the incidence of second breast malignancies overall, although a possible benefit was detected in premenopausal women. These studies, particularly the post hoc analyses, are considered exploratory and need to be confirmed.

Adult↗

PubMed2000

Cancer surveillance series: changing patterns of cutaneous malignant melanoma mortality rates among whites in the United States.

BACKGROUND: Mortality from melanoma among whites is still increasing in the United States. In this study, we describe the changing patterns of melanoma mortality rates among whites by demographic factors and geography and further assess the relationship between the geographic patterns and the UV radiation (UV-B) level. METHODS: Age-adjusted incidence and mortality rates were computed by use of the 1970 U.S. population standard. Annual percent changes of mortality were estimated by fitting regression lines to the logarithm of rates. The relationships between melanoma mortality rates and UV-B level over time were assessed by weighted regressions. All statistical tests were two-sided. RESULTS: From 1950-1954 through 1990-1994, melanoma mortality rates increased by 191% and 84% among males and females, respectively. Mortality rates peaked in the 1930 through 1950 birth cohorts for females and in the 1935 through 1950 birth cohorts for males. In the 1950 through 1969 study period, melanoma mortality rates showed a strong North-South gradient, but the gradient weakened in recent periods. The absolute change in mortality for a 10% increase in UV-B among females decreased from 0.08 additional deaths per 100 000 person-years in 1950-1959 to 0.01 additional deaths in 1990-1995. In contrast, the absolute change in mortality among males showed little change over time; additional deaths increased from 0.11 to 0.12 per 100 000 person-years. CONCLUSIONS: Melanoma mortality in the United States reflects the complex interplay of UV radiation levels in each geographic region, the sun-protection behaviors of each generation of males and females in childhood and adulthood, the geographic mobility of the population, and the risk awareness and early detection.

Adult↗