1995
DOI: 10.1002/sim.4780140806
|View full text |Cite
|
Sign up to set email alerts
|

Logistic regression in case‐control studies: The effect of using independent as dependent variables

Abstract: In case-control studies, cases are sampled separately from controls. In such studies the primary analysis concerns the estimation of the effect of covariables on being a case or a control. To explore causal pathways, further secondary analysis could concern the relationships among the covariables. In this paper the validity of such secondary analysis is addressed. In particular, the use of multiple logistic regression in case-control studies where the dependent variable is not the case/control indicator is exp… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
30
0

Year Published

2001
2001
2016
2016

Publication Types

Select...
7

Relationship

0
7

Authors

Journals

citations
Cited by 29 publications
(30 citation statements)
references
References 8 publications
0
30
0
Order By: Relevance
“…Where neither of the above conditions are satisÿed, Nagelkerke et al [1] advocate ÿtting a standard prospective regression model to Y 2 in terms of x using the Y 1 -control data only. We would expect this to be approximately valid when the proportion of cases in the population is small.…”
Section: Use Only the Controls (Cont)mentioning
confidence: 99%
See 2 more Smart Citations
“…Where neither of the above conditions are satisÿed, Nagelkerke et al [1] advocate ÿtting a standard prospective regression model to Y 2 in terms of x using the Y 1 -control data only. We would expect this to be approximately valid when the proportion of cases in the population is small.…”
Section: Use Only the Controls (Cont)mentioning
confidence: 99%
“…Fitting a standard prospective regression model to Y 2 in terms of x is justiÿed theoretically if Y 1 and Y 2 are conditionally independent given x. If Y 1 and x are conditionally independent given Y 2 , we can still ÿt a logistic regression with only the constant term being a ected by the sampling scheme; see Nagelkerke et al [1]. These two sets of conditions are vaguely interesting but, except where x is discrete with su ciently few categories for the use of saturated models to be feasible, they are of little use for data analysis.…”
Section: Introductionmentioning
confidence: 99%
See 1 more Smart Citation
“…In particular, one cannot use an explanatory variable as a dependent variable in an auxiliary analysis without special precautions (see Nagelkerke et al 1995).…”
Section: Problems To Avoidmentioning
confidence: 99%
“…Standard analysis of the data available, disregarding the original case-control design, can clearly be biased by the over-representation of cases and the association between the new outcome variable and the case-control indicator. Nagelkerke et al [6] show that valid estimates are only obtained under very strict conditions. If the controls in the case-control study are representative of the non-diseased population, then a cross-sectional analysis of the controls can address questions of prevalence of the new outcome and its association with other exposures measured in the original study.…”
Section: Introductionmentioning
confidence: 98%