Stratified Probabilistic Bias Analysis for Body Mass Index-related Exposure Misclassification in Postmenopausal Women

Hailey R. Banack, Andrew Stokes, Matthew P. Fox, Kathleen M. Hovey, Elizabeth M. Cespedes Feliciano, Erin S. LeBlanc, Chloe Bird, Bette J. Caan, Candyce H. Kroenke, Matthew A. Allison, Scott B Going, Linda Snetselaar, Ting Yuan David Cheng, Rowan T. Chlebowski, Marcia L. Stefanick, Michael J. LaMonte, Jean Wactawski-Wende

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

BACKGROUND: There is widespread concern about the use of body mass index (BMI) to define obesity status in postmenopausal women because it may not accurately represent an individual's true obesity status. The objective of the present study is to examine and adjust for exposure misclassification bias from using an indirect measure of obesity (BMI) compared with a direct measure of obesity (percent body fat). METHODS: We used data from postmenopausal non-Hispanic black and non-Hispanic white women in the Women's Health Initiative (n=126,459). Within the Women's Health Initiative, a sample of 11,018 women were invited to participate in a sub-study involving dual-energy x-ray absorptiometry scans. We examined indices of validity comparing BMI-defined obesity (≥30 kg/m), with obesity defined by percent body fat. We then used probabilistic bias analysis models stratified by age and race to explore the effect of exposure misclassification on the obesity-mortality relationship. RESULTS: Validation analyses highlight that using a BMI cutpoint of 30 kg/m to define obesity in postmenopausal women is associated with poor validity. There were notable differences in sensitivity by age and race. Results from the stratified bias analysis demonstrated that failing to adjust for exposure misclassification bias results in attenuated estimates of the obesity-mortality relationship. For example, in non-Hispanic white women 50-59 years of age, the conventional risk difference was 0.017 (95% confidence interval = 0.01, 0.023) and the bias-adjusted risk difference was 0.035 (95% simulation interval = 0.028, 0.043). CONCLUSIONS: These results demonstrate the importance of using quantitative bias analysis techniques to account for nondifferential exposure misclassification of BMI-defined obesity. See video abstract at, http://links.lww.com/EDE/B385.

Original languageEnglish (US)
Pages (from-to)604-613
Number of pages10
JournalEpidemiology (Cambridge, Mass.)
Volume29
Issue number5
DOIs
StatePublished - Sep 1 2018

Fingerprint

Body Mass Index
Obesity
Women's Health
Adipose Tissue
Mortality
X-Rays
Confidence Intervals

ASJC Scopus subject areas

  • Epidemiology

Cite this

Banack, H. R., Stokes, A., Fox, M. P., Hovey, K. M., Cespedes Feliciano, E. M., LeBlanc, E. S., ... Wactawski-Wende, J. (2018). Stratified Probabilistic Bias Analysis for Body Mass Index-related Exposure Misclassification in Postmenopausal Women. Epidemiology (Cambridge, Mass.), 29(5), 604-613. https://doi.org/10.1097/EDE.0000000000000863

Stratified Probabilistic Bias Analysis for Body Mass Index-related Exposure Misclassification in Postmenopausal Women. / Banack, Hailey R.; Stokes, Andrew; Fox, Matthew P.; Hovey, Kathleen M.; Cespedes Feliciano, Elizabeth M.; LeBlanc, Erin S.; Bird, Chloe; Caan, Bette J.; Kroenke, Candyce H.; Allison, Matthew A.; Going, Scott B; Snetselaar, Linda; Cheng, Ting Yuan David; Chlebowski, Rowan T.; Stefanick, Marcia L.; LaMonte, Michael J.; Wactawski-Wende, Jean.

In: Epidemiology (Cambridge, Mass.), Vol. 29, No. 5, 01.09.2018, p. 604-613.

Research output: Contribution to journalArticle

Banack, HR, Stokes, A, Fox, MP, Hovey, KM, Cespedes Feliciano, EM, LeBlanc, ES, Bird, C, Caan, BJ, Kroenke, CH, Allison, MA, Going, SB, Snetselaar, L, Cheng, TYD, Chlebowski, RT, Stefanick, ML, LaMonte, MJ & Wactawski-Wende, J 2018, 'Stratified Probabilistic Bias Analysis for Body Mass Index-related Exposure Misclassification in Postmenopausal Women', Epidemiology (Cambridge, Mass.), vol. 29, no. 5, pp. 604-613. https://doi.org/10.1097/EDE.0000000000000863
Banack, Hailey R. ; Stokes, Andrew ; Fox, Matthew P. ; Hovey, Kathleen M. ; Cespedes Feliciano, Elizabeth M. ; LeBlanc, Erin S. ; Bird, Chloe ; Caan, Bette J. ; Kroenke, Candyce H. ; Allison, Matthew A. ; Going, Scott B ; Snetselaar, Linda ; Cheng, Ting Yuan David ; Chlebowski, Rowan T. ; Stefanick, Marcia L. ; LaMonte, Michael J. ; Wactawski-Wende, Jean. / Stratified Probabilistic Bias Analysis for Body Mass Index-related Exposure Misclassification in Postmenopausal Women. In: Epidemiology (Cambridge, Mass.). 2018 ; Vol. 29, No. 5. pp. 604-613.
@article{2b8418e1b1eb4282bdbc32bd5e28a6d5,
title = "Stratified Probabilistic Bias Analysis for Body Mass Index-related Exposure Misclassification in Postmenopausal Women",
abstract = "BACKGROUND: There is widespread concern about the use of body mass index (BMI) to define obesity status in postmenopausal women because it may not accurately represent an individual's true obesity status. The objective of the present study is to examine and adjust for exposure misclassification bias from using an indirect measure of obesity (BMI) compared with a direct measure of obesity (percent body fat). METHODS: We used data from postmenopausal non-Hispanic black and non-Hispanic white women in the Women's Health Initiative (n=126,459). Within the Women's Health Initiative, a sample of 11,018 women were invited to participate in a sub-study involving dual-energy x-ray absorptiometry scans. We examined indices of validity comparing BMI-defined obesity (≥30 kg/m), with obesity defined by percent body fat. We then used probabilistic bias analysis models stratified by age and race to explore the effect of exposure misclassification on the obesity-mortality relationship. RESULTS: Validation analyses highlight that using a BMI cutpoint of 30 kg/m to define obesity in postmenopausal women is associated with poor validity. There were notable differences in sensitivity by age and race. Results from the stratified bias analysis demonstrated that failing to adjust for exposure misclassification bias results in attenuated estimates of the obesity-mortality relationship. For example, in non-Hispanic white women 50-59 years of age, the conventional risk difference was 0.017 (95{\%} confidence interval = 0.01, 0.023) and the bias-adjusted risk difference was 0.035 (95{\%} simulation interval = 0.028, 0.043). CONCLUSIONS: These results demonstrate the importance of using quantitative bias analysis techniques to account for nondifferential exposure misclassification of BMI-defined obesity. See video abstract at, http://links.lww.com/EDE/B385.",
author = "Banack, {Hailey R.} and Andrew Stokes and Fox, {Matthew P.} and Hovey, {Kathleen M.} and {Cespedes Feliciano}, {Elizabeth M.} and LeBlanc, {Erin S.} and Chloe Bird and Caan, {Bette J.} and Kroenke, {Candyce H.} and Allison, {Matthew A.} and Going, {Scott B} and Linda Snetselaar and Cheng, {Ting Yuan David} and Chlebowski, {Rowan T.} and Stefanick, {Marcia L.} and LaMonte, {Michael J.} and Jean Wactawski-Wende",
year = "2018",
month = "9",
day = "1",
doi = "10.1097/EDE.0000000000000863",
language = "English (US)",
volume = "29",
pages = "604--613",
journal = "Epidemiology",
issn = "1044-3983",
publisher = "Lippincott Williams and Wilkins",
number = "5",

}

TY - JOUR

T1 - Stratified Probabilistic Bias Analysis for Body Mass Index-related Exposure Misclassification in Postmenopausal Women

AU - Banack, Hailey R.

AU - Stokes, Andrew

AU - Fox, Matthew P.

AU - Hovey, Kathleen M.

AU - Cespedes Feliciano, Elizabeth M.

AU - LeBlanc, Erin S.

AU - Bird, Chloe

AU - Caan, Bette J.

AU - Kroenke, Candyce H.

AU - Allison, Matthew A.

AU - Going, Scott B

AU - Snetselaar, Linda

AU - Cheng, Ting Yuan David

AU - Chlebowski, Rowan T.

AU - Stefanick, Marcia L.

AU - LaMonte, Michael J.

AU - Wactawski-Wende, Jean

PY - 2018/9/1

Y1 - 2018/9/1

N2 - BACKGROUND: There is widespread concern about the use of body mass index (BMI) to define obesity status in postmenopausal women because it may not accurately represent an individual's true obesity status. The objective of the present study is to examine and adjust for exposure misclassification bias from using an indirect measure of obesity (BMI) compared with a direct measure of obesity (percent body fat). METHODS: We used data from postmenopausal non-Hispanic black and non-Hispanic white women in the Women's Health Initiative (n=126,459). Within the Women's Health Initiative, a sample of 11,018 women were invited to participate in a sub-study involving dual-energy x-ray absorptiometry scans. We examined indices of validity comparing BMI-defined obesity (≥30 kg/m), with obesity defined by percent body fat. We then used probabilistic bias analysis models stratified by age and race to explore the effect of exposure misclassification on the obesity-mortality relationship. RESULTS: Validation analyses highlight that using a BMI cutpoint of 30 kg/m to define obesity in postmenopausal women is associated with poor validity. There were notable differences in sensitivity by age and race. Results from the stratified bias analysis demonstrated that failing to adjust for exposure misclassification bias results in attenuated estimates of the obesity-mortality relationship. For example, in non-Hispanic white women 50-59 years of age, the conventional risk difference was 0.017 (95% confidence interval = 0.01, 0.023) and the bias-adjusted risk difference was 0.035 (95% simulation interval = 0.028, 0.043). CONCLUSIONS: These results demonstrate the importance of using quantitative bias analysis techniques to account for nondifferential exposure misclassification of BMI-defined obesity. See video abstract at, http://links.lww.com/EDE/B385.

AB - BACKGROUND: There is widespread concern about the use of body mass index (BMI) to define obesity status in postmenopausal women because it may not accurately represent an individual's true obesity status. The objective of the present study is to examine and adjust for exposure misclassification bias from using an indirect measure of obesity (BMI) compared with a direct measure of obesity (percent body fat). METHODS: We used data from postmenopausal non-Hispanic black and non-Hispanic white women in the Women's Health Initiative (n=126,459). Within the Women's Health Initiative, a sample of 11,018 women were invited to participate in a sub-study involving dual-energy x-ray absorptiometry scans. We examined indices of validity comparing BMI-defined obesity (≥30 kg/m), with obesity defined by percent body fat. We then used probabilistic bias analysis models stratified by age and race to explore the effect of exposure misclassification on the obesity-mortality relationship. RESULTS: Validation analyses highlight that using a BMI cutpoint of 30 kg/m to define obesity in postmenopausal women is associated with poor validity. There were notable differences in sensitivity by age and race. Results from the stratified bias analysis demonstrated that failing to adjust for exposure misclassification bias results in attenuated estimates of the obesity-mortality relationship. For example, in non-Hispanic white women 50-59 years of age, the conventional risk difference was 0.017 (95% confidence interval = 0.01, 0.023) and the bias-adjusted risk difference was 0.035 (95% simulation interval = 0.028, 0.043). CONCLUSIONS: These results demonstrate the importance of using quantitative bias analysis techniques to account for nondifferential exposure misclassification of BMI-defined obesity. See video abstract at, http://links.lww.com/EDE/B385.

UR - http://www.scopus.com/inward/record.url?scp=85056346070&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85056346070&partnerID=8YFLogxK

U2 - 10.1097/EDE.0000000000000863

DO - 10.1097/EDE.0000000000000863

M3 - Article

C2 - 29864084

AN - SCOPUS:85056346070

VL - 29

SP - 604

EP - 613

JO - Epidemiology

JF - Epidemiology

SN - 1044-3983

IS - 5

ER -