Search PubMed⌕ Search

Biomedical subjects

D Pregibon

Publications and source records attributed to D Pregibon.

7 recordsLinked to original sources
PubMed1988

Neuroendocrine aspects of primary endogenous depression. VII. Logistic regression analysis of matched patient-control hormone data for discrimination between groups.

We previously reported significant alterations in hypothalamo-pituitary-adrenal cortical (HPA) and thyroid (HPT) axis function in a group of 40 primary, RDC definite endogenous depressives compared to 40 individually matched normal control subjects. In those reports, univariate statistical tests for matched samples were used (paired t-tests). We next wished to use both the HPA and the HPT measures in a multivariate analysis to discriminate the patients from the controls, but the paired nature of the hormone data obviated the use of standard techniques. With an adaptation of the BMDP PLR statistical program, we performed a matched logistic regression analysis, in stepwise fashion, which resulted in a model incorporating two HPA variables and one HPT variable. From this model, odds ratios which associate being depressed or not with levels of the HPA and HPT measures were estimated. The median odds ratio for the 40 matched pairs was 7.8. To illustrate the model, we present graphs of estimated odds ratios of patients with selected hormone values compared to a typical normal subject with hormone values close to the median values of the 40 controls in our study.

Adult↗
PubMed1987

A quantitative analysis of the effects of activity and time of day on the diurnal variations of blood pressure.

The effects of activity and time of day on blood pressure (BP) were analyzed in 461 patients with untreated hypertension who wore a noninvasive portable BP recorder which took readings every 15 minutes for 24 hours. Patients recorded activity and location in a diary. The data were analyzed separately for two groups of patients: the 190 who stayed at home and the 271 who went to work. The effects of 16 different activities on BP were estimated by relating the BP to the associated activity and to the individual's clinic BP. Blood pressure was higher at work than at home, but the increment of BP for individual activities was similar in the two locations. The overall effect of activities on BP variability was computed using a one-way analysis of covariance model. For the patients who went to work this model accounted for 40% of the observed variation (R2) for systolic and 39% for diastolic BP. A similar model using time of day instead of activity accounted for 33% of variability in both systolic and diastolic BP. Combining activity and time of day was little better than activity alone (41% for both). After allowing for the effects of activity on BP, where sleep is one of the activities, there was no significant diurnal variation of BP. We conclude that there is no important circadian rhythm of BP which is independent of activity.

Adolescent↗
PubMed1986

Behavioral determinants of 24-hour blood pressure patterns in borderline hypertension.

Borderline hypertension is a heterogeneous condition; only a minority of patients will progress to fixed essential hypertension or suffer cardiovascular damage. Ambulatory blood pressure monitoring can detect the subgroup of borderlines with sustained blood pressure elevation outside the office setting and can also provide a measure of blood pressure variability. Patients in the former group may be at greatest risk for cardiovascular complications while the latter have been theorized to be at risk for progression to fixed hypertension. Evidence we have gathered utilizing ambulatory monitoring does not support the contention that there is a subgroup of borderline hypertensives with excessive blood pressure variability in natural settings. Such recordings have been of benefit in identifying patients with excessive pressor responses to office visits. Future studies employing ambulatory monitoring may be useful in the detection of clinically important subgroups of borderline hypertension.

Behavior↗
PubMed1984

Data analytic methods for matched case-control studies.

The recent introduction of complex multivariate statistical models in matched case-control studies is a mixed blessing. Their use can lead to a better understanding of the way in which many variables contribute to the risk of disease. On the other hand, these powerful methods can obscure salient features in the data that might have been detected by other, less sophisticated methods. This shortcoming is due to a lack of support methodology for the routine use of these models. Satisfactory computation of estimated relative risks and their standard errors is not sufficient justification for the fitted model. Goodness of fit must be examined if inferences are to be trusted. This paper is concerned with the analysis of matched case-control studies with logistic models. Analogies of these models to linear regression models are emphasized. In particular, basic concepts such as analysis of variance, multiple correlation coefficient, one-degree-of-freedom tests, and residual analysis are discussed. The fairly new field of regression diagnostics is also introduced. All procedures are illustrated on a study of bladder cancer in males.

Analysis of Variance↗
PubMed1982

Resistant fits for some commonly used logistic models with medical application.

Logistic regression-type models are used in many applications. Some examples include the classical dose-response experiment, prospective and retrospective studies of disease incidence (with and without matching), and the analysis of ordinal data. In most instances, the model is fitted by the method of maximum likelihood, which, like least squares, is sensitive to atypical observations. An alternative to maximum likelihood is proposed and illustrated by examples.

Epidemiologic Methods↗