Corrections for exposure measurement error in logistic regression models with an application to nutritional data

Kuha, Jouni

doi:10.1002/sim.4780131105

Cited by 58 publications

(60 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Measurement errors in multicentre studies are caused by variation within and between cohorts and calibration of subsamples can be used to control for this variation. 23,24 Repeated measurement is another principle to reduce the variance within individuals but this will apply only if there is no correlated error. 25,26 Most models used in nutritional epidemiology assume no correlated error.…”

Section: (Lower Part) As a Function Of Validity Of Measurements And Cmentioning

confidence: 99%

Bias in diet assessment methods--consequences of collinearity and measurement errors on power and observed relative risks

Elmståhl

Gullberg²

1997

International Journal of Epidemiology

View full text Add to dashboard Cite

Several ongoing cohort studies in Europe and US use different dietary assessment methods to increase our knowledge about diet and disease relationships concerning cardiovascular disease, osteoporosis and cancer. Diet history methods and food frequency questionnaires are commonly used methods to measure usual food intake in case-control studies and cohort studies. However, misclassification imposed by the methods limit the possibilities of collecting reliable data on dietary intake. The validity of the dietary methods depends on several components, like accurate estimations of frequency and portion sizes, the quality of food composition data bases used, and data collection procedure; all components which are crucial for interpretation of the dietary results. If the food list does not include commonly consumed food items in the chosen study population, or dietary methods with predefined questions are used, the sample characteristics must be homogeneous for several factors like gender, age, ethnicity and social class. This might introduce narrower distribution of potential explanatory variables. Non-representative study samples will also restrict valid estimations of populations at risk. This paper intends to evaluate some methodological issues of logistic models regarding misclassification and attenuation of estimated observed relative risks.Different diet assessment methods are used in some of the ongoing cohort studies and these methods have been validated in the past. A need for further development of the present dietary methods was put forward at an international meeting on diet assessment methods. Studies using semi-quantitative food frequency questionnaires were also recommended to include standardized substudies with recall or record methods to get quantitative calibrated estimates of nutrient intakes.2 Other methods, for example meal-based methods like dietary record, diet histories and 24-hour recalls are more timeconsuming but they have some advantages since they cover the consumption of all foods.3 Although, a gold standard is lacking for validation of the different diet assessment methods, comparison of the method's Bias in Diet Assessment MethodsConsequences of Collinearity and Measurement Errors on Power and Observed Relative Risks SÖLVE ELMSTÅHL AND BO GULLBERGElmståhl S (Department of Community Medicine, Lund University, S-205 02 Malmö, Sweden) and Gullberg B. Bias in diet assessment methods-consequences of collinearity and measurement errors on power and observed relative risks. International Journal of Epidemiology 1997; 26: 1071-1079. Background. If several risk factors for disease are considered in a regression model and these factors are affected by measurement errors, the observed relative risk will be attenuated. In nutritional epidemiology, several nutrient variables show strong correlation, described as collinearity. The observed relative risk will then depend not only on the validity of the chosen diet assessment method but also on collinearity between variables in the model. Methods. Th...

show abstract

Section: (Lower Part) As a Function Of Validity Of Measurements And Cmentioning

confidence: 99%

Bias in diet assessment methods--consequences of collinearity and measurement errors on power and observed relative risks

Elmståhl

Gullberg²

1997

International Journal of Epidemiology

View full text Add to dashboard Cite

show abstract

“…The Taylor series approximation for the standard error for the linear approximation approach will always result in a larger variance. Kuha (1994) extended the linear approximation by improving the approxmation of the logistic model using a second-order Taylor series instead of the first-order series.…”

Section: Regression Calibration Methodsmentioning

confidence: 99%

A Bayesian Adjustment for Covariate Misclassification with Correlated Binary Outcome Data

Ren

Stone

2007

Journal of Applied Statistics

View full text Add to dashboard Cite

“…Generalizing Kuha's derivation (Kuha 1994) to the heteroscedastic measurement error variance case, by using a second-order Taylor series expansion of logit [Pr(Y=1)] with respect to 1 β around 1 0 β = under the rare disease assumption and with | , x X U normally distributed with linear mean and variance, ( | , ;…”

Section: Iterative Methodsmentioning

confidence: 99%

“…(Kuha 1994 As is evident from (6), this estimator can only be used for models with scalar x. For multivariate x with heteroscedastic covariance for x | X,U , a term is added to the model for E(Y|X,U) equal to…”

Section: The First Row Of γ Denoted ˆˆ( )mentioning

confidence: 99%

See 1 more Smart Citation

Regression Calibration with Heteroscedastic Error Variance

Spiegelman

Logan²,

Grove³

2011

The International Journal of Biostatistics

View full text Add to dashboard Cite

The problem of covariate measurement error with heteroscedastic measurement error variance is considered. Standard regression calibration assumes that the measurement error has a homoscedastic measurement error variance. An estimator is proposed to correct regression coefficients for covariate measurement error with heteroscedastic variance. Point and interval estimates are derived. Validation data containing the gold standard must be available. This estimator is a closed-form correction of the uncorrected primary regression coefficients, which may be of logistic or Cox proportional hazards model form, and is closely related to the version of regression calibration developed by Rosner et al. (1990). The primary regression model can include multiple covariates measured without error. The use of these estimators is illustrated in two data sets, one taken from occupational epidemiology (the ACE study) and one taken from nutritional epidemiology (the Nurses' Health Study). In both cases, although there was evidence of moderate heteroscedasticity, there was little difference in estimation or inference using this new procedure compared to standard regression calibration. It is shown theoretically that unless the relative risk is large or measurement error severe, standard regression calibration approximations will typically be adequate, even with moderate heteroscedasticity in the measurement error model variance. In a detailed simulation study, standard regression calibration performed either as well as or better than the new estimator. When the disease is rare and the errors normally distributed, or when measurement error is moderate, standard regression calibration remains the method of choice.

show abstract

Corrections for exposure measurement error in logistic regression models with an application to nutritional data

Cited by 58 publications

References 17 publications

Bias in diet assessment methods--consequences of collinearity and measurement errors on power and observed relative risks

Bias in diet assessment methods--consequences of collinearity and measurement errors on power and observed relative risks

A Bayesian Adjustment for Covariate Misclassification with Correlated Binary Outcome Data

Regression Calibration with Heteroscedastic Error Variance

Contact Info

Product

Resources

About