Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Measurement error models for ordinal exposure variables measured with error

B A Rosner1

  • 1Channing Laboratory, Brigham and Women's Hospital, Boston, MA 02115, USA.

Statistics in Medicine
|February 15, 1996
PubMed
Summary

This study introduces a measurement error model to correct nutrient intake estimates in dietary epidemiology. The model improves relative risk assessments for diseases like breast cancer, accounting for nutrient measurement errors.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Association of oral contraceptives and tubal ligation with risk of early natural menopause.

Human reproduction (Oxford, England)·2021
Same author

Associations of early life and adulthood adiposity with risk of epithelial ovarian cancer.

Annals of oncology : official journal of the European Society for Medical Oncology·2018
Same author

Adult adiposity and risk of early menopause.

Human reproduction (Oxford, England)·2017
Same author

Validation of Rosner-Colditz breast cancer incidence model using an independent data set, the California Teachers Study.

Breast cancer research and treatment·2013
Same author

Comparison of aspects of smoking among the four histological types of lung cancer.

Tobacco control·2008
Same author

A prospective study of dietary carbohydrate quantity and quality in relation to risk of ovulatory infertility.

European journal of clinical nutrition·2007

Area of Science:

  • Dietary Epidemiology
  • Biostatistics
  • Nutritional Science

Background:

  • Nutrient variables in dietary epidemiology are frequently analyzed using quintile scales.
  • Measurement error in nutrient assessment is a significant challenge, potentially biasing relative risk estimates.
  • Accurate risk assessment requires addressing these measurement errors.

Purpose of the Study:

  • To propose a novel measurement error model for nutrient intake expressed in quintiles.
  • To correct relative risk estimates by accounting for errors in nutrient measurement.
  • To apply the model to assess breast cancer risk associated with fat and alcohol intake.

Main Methods:

  • Developed a measurement error model relating true (diet record) and noisy (food frequency) nutrient values in quintiles.

Related Experiment Videos

  • Estimated model parameters using validation study data.
  • Applied the validated model to epidemiological data from the Nurses' Health Study.

    • Main Results:

    • The proposed model effectively relates true and noisy nutrient values within the quintile scale.
    • Estimated corrected relative risks for breast cancer risk factors.
    • Demonstrated the model's utility in a real-world epidemiological dataset.

    Conclusions:

    • The developed measurement error model provides a robust method for correcting nutrient intake data in dietary epidemiology.
    • This approach enhances the accuracy of relative risk estimations, particularly for chronic diseases.
    • The findings support the use of measurement error correction in nutritional research to improve health outcome assessments.