Estimation of the multivariate logistic risk function: A comparison of the discriminant function and maximum likelihood approaches

Halperin, Max; Blackwelder, William C.; Verter, Joel

doi:10.1016/0021-9681(71)90106-8

Cited by 222 publications

(63 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…'9 is a linear regression which uses the logarithm of the relative proportion of cases to non-cases as the dependent variable. It is closely related to discriminant analysis, and may be regarded as an extension of discriminant analysis to a broader class of models.20 21 The regression coefficient for each independent variable provides an estimate of the "odds ratio," the ratio between the chance an event will occur when a given factor is present to the chance it will occur when the factor is absent. For example, an odds ratio of 4.95 for the risk factor "No prenatal care" with respect to birthweight less than 2501 grams can be interpreted to mean that if all other factors are the same a woman who has not received prenatal care is about five times as likely to give birth to an infant weighing less than 2501 grams as is a woman who has received some prenatal care.…”

Section: Methodsmentioning

confidence: 99%

The risk of low birthweight.

Eisner¹,

Brazie²,

Pratt³

et al. 1979

Am J Public Health

150

View full text Add to dashboard Cite

Descriptive analyses of birthweight for single live births in the United States during 1974, using birth certificate information, show that several factors are associated with a high incidence of low birthweight babies. Multivariate analyses were performed to determine correlates of low birthweight. When other factors are held constant, race not white, previous reproductive loss, short interpregnancy interval, outof-wedlock birth, no prenatal care, and maternal age Recent studies of pregnancy outcome have shown a progressive decline of perinatal and infant mortality in all risk categories, including mortality for small preterm infants.' The decline has come at a time of change in many aspects of maternal and infant care: associated factors include improved perinatal services available to both mother and baby, improved care or prevention of specific diseases, and an increased availability of means for spacing or preventing pregnancies.In our study of birth certificate information to evaluate specific factors which increase the risk of a poor pregnancy outcome, we have followed the suggestion of Lewis, et al.,2 and use birthweight distribution as an indicator of pregnancy outcome. Although infant mortality rates remain the primary indicator, low birthweight is also an indicator of considerable interest. Low birthweight is associated with increased perinatal and infant mortality and morbidity, including adverse sequelae such as mental retardation and learning disabilities.Many previous studies have shown that birthweight, in- The study from which the present report has been drawn'4 comprised an examination of United States birth registration data for single live births in 1974. For each race it related the proportion of births which were low birthweight (less than 2501 grams) to risk factors ascertainable from birth certificates. We summarize here the principal findings of that study and report the results of multivariate analyses which examine the risk of having a low birthweight baby by race, maternal age, wedlock status, prenatal care, and maternal education for primigravida, and by these factors and birth order, reproductive history, and interpregnancy interval for multigravida.

show abstract

Section: Methodsmentioning

confidence: 99%

The risk of low birthweight.

Eisner¹,

Brazie²,

Pratt³

et al. 1979

Am J Public Health

150

View full text Add to dashboard Cite

show abstract

“…commun.). However, in comparison with discriminant fiinction analysis, Brennan et al (1996), Halperin et al (1971), and Press and Wilson (1978) achieved greater group separation (between groups of independent variables associated with different values of the dependent variable) with LR.…”

Section: Analysesmentioning

confidence: 97%

Distribution, Abundance, and Habitat Characteristics of the Buff-Breasted Flycatcher in Arizona

Martin¹

1999

The Condor

View full text Add to dashboard Cite

INFORMATION TO USERSThis manuscript has been reproduced from the microfilm master. UMI films the text directfy fix>m the originai or copy submitted. Thus, some thesis and dissertation copies are in ^ewiiter &ce; vt^e others may be fi'om aiQr Qrpe of computer printer.The quality of this reprodnctioii is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affea reproduction.In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.Oversize materials (e.g., nuq>s, drawings, charts) are reproduced by sectioning the original, b^inning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

show abstract

“…Figure 7 gives the expected frequency distribution of plasma cholesterol levels for each of the six apo E phenotypes estimated from the Ottawa study 54 and the risk curve for CAD associated with cholesterol estimated from the Framingham experience. 128 The area under each curve gives the estimated proportion of individuals in the Ottawa population with the corresponding phenotype. The expected risk associated with the average cholesterol level of each phenotype is denoted on the risk curve.…”

Section: Impact Of Apo E On Riskmentioning

confidence: 99%

“…14 used the logistic risk equation from the Framingham study 128 and the Ottawa data 54 presented in Figure 7 to estimate the average risk of CAD associated wtth the e2, e3, and e4 alleles. Their calculations suggest that the subpopulation carrying the e4 allele has an average risk 1.43 times as great as that for the group carrying the e2 allele.…”

Section: Impact Of Apo E On Riskmentioning

confidence: 99%

Apolipoprotein E polymorphism and atherosclerosis.

1988

View full text Add to dashboard Cite

Lipoproteins play a central role in the development of atherosclerotic cardiovascular disease in humans. The plasma concentrations of lipoproteins and their metabolic fates are modulated by apolipoproteins on the surface of these lipid-rich particles. The hypothesis has been raised that genetic variation in apolipoproteins is a major determinant of the interindividual variation in susceptibility to coronary artery disease (CAD). In humans, the structural gene locus for plasma apolipoprotein E (apo E) is polymorphic:1 " 6 three common alleles, designated e4, e3, and e2, code for three major apo E isoforms in plasma, respectively designated apo E4, apo E3, and apo E2. Apo E is a key protein in the modulation of the metabolism of the highly atherogenic apo B-containing lipoproteins. Therefore, apo E provides a good model to test the above hypothesis.The existence of a polymorphism for such an important apolipoprotein has been a source of major interest in recent years and has raised many questions for clinicians, epidemiologists, geneticists, and biochemists. The similarities of the relative frequencies of the three common alleles among populations representing very different ethnic, cultural, and geographic settings; their different effects on •lipoprotein metabolism; and their association with certain lipid transport disorders have stimulated research aimed at a better understanding of their role in health and disease. A wealth of information 7 " 18 has accrued from studies on apolipoprotein E structure, metabolism, and function. These have considerably advanced our knowledge of the role of apo E in normal and abnormal lipoprotein metabolism, the complex interactions leading to hyperfipidemia and atherosclerosis, and the pathogenesis of type III hyperiipoproteinemia. Of major interest is the observation that the protein products of the e4 and E2 alleles have separate, and in some circumstances opposite, influences on plasma lipo- protein concentrations, a finding which leads to the hypothesis that the apo E polymorphism ranks among the major factors involved in determining interindividual differences in the initiation and progression of atherosclerosis. The evidence associating the apo E polymorphism with altered lipid and lipoprotein levels and atherosclerosis will be reviewed here. First, the apo E gene and gene product will be considered with a brief overview of molecular biology, protein chemistry, and methods for phenotyping. Second, the impact of variation at the apo E locus on plasma lipids and lipoproteins in health and disease will be discussed. Third, the metabolic interactions accounting for the various effects of the different alleles on lipids and lipoproteins will be considered to provide a unifying interpretation as to how apo E modulates the metabolism of apo B-containing lipoproteins and their remnants. Last, direct evidence associating the apo E polymorphism and atherosclerosis will be reviewed and discussed. This review is not intended to be all-inclusive, and the reader is referred to recent r...

show abstract

Estimation of the multivariate logistic risk function: A comparison of the discriminant function and maximum likelihood approaches

Cited by 222 publications

References 3 publications

The risk of low birthweight.

The risk of low birthweight.

Distribution, Abundance, and Habitat Characteristics of the Buff-Breasted Flycatcher in Arizona

Apolipoprotein E polymorphism and atherosclerosis.

Contact Info

Product

Resources

About