Correcting for multivariate measurement error by regression calibration in meta-analyses of epidemiological studies

Angela Mary Wood; S. G. Thompson; J. B. Kostis; A. C. Wilson; K. Wu; M. Benderly; U. Goldbourt; J. Willeit; S. Kiechl; J. W.G. Yarnell; P. M. Sweetman; P. C. Elwood; M. Cushman; R. P. Tracy; A. Tybjærg-Hansen; F. Haverkate; A. J. Lee; F. B. Smith; V. Salomaa; K. Harald; V. Rasi; P. Jousilahti; J. Pekkanen; R. D'Agostino; P. W.F. Wilson; G. Tofler; D. Levy; R. Marchioli; F. Valagussa; A. Rosengren; G. Lappas; H. Eriksson; P. Cremer; D. Nagel; J. D. Curb; B. Rodriguez; K. Yano; J. T. Salonen; K. Nyyssönen; T. P. Tuomainen; B. Hedblad; G. Engström; G. Berglund; H. Loewel; H. W. Hense; T. W. Mead; J. A. Cooper; B. De Stavola; C. Knottenbelt; G. J. Miller; J. A. Cooper; K. A. Bauer; R. D. Rosenberg; S. Sato; A. Kitamura; Y. Naito; H. Iso; K. Harald; V. Rasi; E. Vahtera; P. Jousilahti; T. Palosuo; P. Ducimetriere; P. Amouyel; D. Arveiler; A. E. Evans; J. Ferrieres; I. Juhan-Vague; A. Bingham; H. Schulte; G. Assmann; B. Cantin; B. Lamarche; J. P. Després; H. Tunstall-Pedoe; G. D.O. Lowe; M. Woodward; Y. Ben-Shlomo; G. Davey Smith; V. Palmieri; J. L. Yeh; T. W. Meade; P. Brennan; C. Knottenbelt; P. Ridker; F. Rodeghiero; A. Tosetto; J. Shepherd; G. D.O. Lowe; I. Ford; M. Robertson; E. Brunner; M. Shipley; E. J.M. Feskens; D. Kromhout; E. Di Angelantonio; S. Kaptoge; S. Lewington; G. D.O. Lowe

doi:10.1002/sim.3530

Correcting for multivariate measurement error by regression calibration in meta-analyses of epidemiological studies

Angela Mary Wood^*, S. G. Thompson, J. B. Kostis, A. C. Wilson, K. Wu, M. Benderly, U. Goldbourt, J. Willeit, S. Kiechl, J. W.G. Yarnell, P. M. Sweetman, P. C. Elwood, M. Cushman, R. P. Tracy, A. Tybjærg-Hansen, F. Haverkate, A. J. Lee, F. B. Smith, V. Salomaa, K. HaraldV. Rasi, P. Jousilahti, J. Pekkanen, R. D'Agostino, P. W.F. Wilson, G. Tofler, D. Levy, R. Marchioli, F. Valagussa, A. Rosengren, G. Lappas, H. Eriksson, P. Cremer, D. Nagel, J. D. Curb, B. Rodriguez, K. Yano, J. T. Salonen, K. Nyyssönen, T. P. Tuomainen, B. Hedblad, G. Engström, G. Berglund, H. Loewel, H. W. Hense, T. W. Mead, J. A. Cooper, B. De Stavola, C. Knottenbelt, G. J. Miller, J. A. Cooper, K. A. Bauer, R. D. Rosenberg, S. Sato, A. Kitamura, Y. Naito, H. Iso, K. Harald, V. Rasi, E. Vahtera, P. Jousilahti, T. Palosuo, P. Ducimetriere, P. Amouyel, D. Arveiler, A. E. Evans, J. Ferrieres, I. Juhan-Vague, A. Bingham, H. Schulte, G. Assmann, B. Cantin, B. Lamarche, J. P. Després, H. Tunstall-Pedoe, G. D.O. Lowe, M. Woodward, Y. Ben-Shlomo, G. Davey Smith, V. Palmieri, J. L. Yeh, T. W. Meade, P. Brennan, C. Knottenbelt, P. Ridker, F. Rodeghiero, A. Tosetto, J. Shepherd, G. D.O. Lowe, I. Ford, M. Robertson, E. Brunner, M. Shipley, E. J.M. Feskens, D. Kromhout, E. Di Angelantonio, S. Kaptoge, S. Lewington, G. D.O. Lowe

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

58 Scopus citations

Abstract

Within-person variability in measured values of multiple risk factors can bias their associations with disease. The multivariate regression calibration (RC) approach can correct for such measurement error and has been applied to studies in which true values or independent repeat measurements of the risk factors are observed on a subsample. We extend the multivariate RC techniques to a meta-analysis framework where multiple studies provide independent repeat measurements and information on disease outcome. We consider the cases where some or all studies have repeat measurements, and compare study-specific, averaged and empirical Bayes estimates of RC parameters. Additionally, we allow for binary covariates (e.g. smoking status) and for uncertainty and time trends in the measurement error corrections. Our methods are illustrated using a subset of individual participant data from prospective long-term studies in the Fibrinogen Studies Collaboration to assess the relationship between usual levels of plasma fibrinogen and the risk of coronary heart disease, allowing for measurement error in plasma fibrinogen and several confounders.

Original language	English
Pages (from-to)	1067-1092
Number of pages	26
Journal	Statistics in Medicine
Volume	28
Issue number	7
DOIs	https://doi.org/10.1002/sim.3530
State	Published - 30 Mar 2009

Funding

Funders	Funder number
National Heart, Lung, and Blood Institute	U01HL080295

Keywords

Individual participant data
Measurement error
Meta-analysis
Regression calibration
Within-person variation

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1002/sim.3530

Cite this

Wood, A. M., Thompson, S. G., Kostis, J. B., Wilson, A. C., Wu, K., Benderly, M., Goldbourt, U., Willeit, J., Kiechl, S., Yarnell, J. W. G., Sweetman, P. M., Elwood, P. C., Cushman, M., Tracy, R. P., Tybjærg-Hansen, A., Haverkate, F., Lee, A. J., Smith, F. B., Salomaa, V., ... Lowe, G. D. O. (2009). Correcting for multivariate measurement error by regression calibration in meta-analyses of epidemiological studies. Statistics in Medicine, 28(7), 1067-1092. https://doi.org/10.1002/sim.3530

@article{7345da42737141f09ec56878e21d7a71,

title = "Correcting for multivariate measurement error by regression calibration in meta-analyses of epidemiological studies",

abstract = "Within-person variability in measured values of multiple risk factors can bias their associations with disease. The multivariate regression calibration (RC) approach can correct for such measurement error and has been applied to studies in which true values or independent repeat measurements of the risk factors are observed on a subsample. We extend the multivariate RC techniques to a meta-analysis framework where multiple studies provide independent repeat measurements and information on disease outcome. We consider the cases where some or all studies have repeat measurements, and compare study-specific, averaged and empirical Bayes estimates of RC parameters. Additionally, we allow for binary covariates (e.g. smoking status) and for uncertainty and time trends in the measurement error corrections. Our methods are illustrated using a subset of individual participant data from prospective long-term studies in the Fibrinogen Studies Collaboration to assess the relationship between usual levels of plasma fibrinogen and the risk of coronary heart disease, allowing for measurement error in plasma fibrinogen and several confounders.",

keywords = "Individual participant data, Measurement error, Meta-analysis, Regression calibration, Within-person variation",

author = "Wood, {Angela Mary} and Thompson, {S. G.} and Kostis, {J. B.} and Wilson, {A. C.} and K. Wu and M. Benderly and U. Goldbourt and J. Willeit and S. Kiechl and Yarnell, {J. W.G.} and Sweetman, {P. M.} and Elwood, {P. C.} and M. Cushman and Tracy, {R. P.} and A. Tybj{\ae}rg-Hansen and F. Haverkate and Lee, {A. J.} and Smith, {F. B.} and V. Salomaa and K. Harald and V. Rasi and P. Jousilahti and J. Pekkanen and R. D'Agostino and Wilson, {P. W.F.} and G. Tofler and D. Levy and R. Marchioli and F. Valagussa and A. Rosengren and G. Lappas and H. Eriksson and P. Cremer and D. Nagel and Curb, {J. D.} and B. Rodriguez and K. Yano and Salonen, {J. T.} and K. Nyyss{\"o}nen and Tuomainen, {T. P.} and B. Hedblad and G. Engstr{\"o}m and G. Berglund and H. Loewel and Hense, {H. W.} and Mead, {T. W.} and Cooper, {J. A.} and {De Stavola}, B. and C. Knottenbelt and Miller, {G. J.} and Cooper, {J. A.} and Bauer, {K. A.} and Rosenberg, {R. D.} and S. Sato and A. Kitamura and Y. Naito and H. Iso and K. Harald and V. Rasi and E. Vahtera and P. Jousilahti and T. Palosuo and P. Ducimetriere and P. Amouyel and D. Arveiler and Evans, {A. E.} and J. Ferrieres and I. Juhan-Vague and A. Bingham and H. Schulte and G. Assmann and B. Cantin and B. Lamarche and Despr{\'e}s, {J. P.} and H. Tunstall-Pedoe and Lowe, {G. D.O.} and M. Woodward and Y. Ben-Shlomo and Smith, {G. Davey} and V. Palmieri and Yeh, {J. L.} and Meade, {T. W.} and P. Brennan and C. Knottenbelt and P. Ridker and F. Rodeghiero and A. Tosetto and J. Shepherd and Lowe, {G. D.O.} and I. Ford and M. Robertson and E. Brunner and M. Shipley and Feskens, {E. J.M.} and D. Kromhout and {Di Angelantonio}, E. and S. Kaptoge and S. Lewington and Lowe, {G. D.O.}",

year = "2009",

month = mar,

day = "30",

doi = "10.1002/sim.3530",

language = "אנגלית",

volume = "28",

pages = "1067--1092",

journal = "Statistics in Medicine",

issn = "0277-6715",

publisher = "John Wiley and Sons Ltd",

number = "7",

}

Wood, AM, Thompson, SG, Kostis, JB, Wilson, AC, Wu, K, Benderly, M , Goldbourt, U, Willeit, J, Kiechl, S, Yarnell, JWG, Sweetman, PM, Elwood, PC, Cushman, M, Tracy, RP, Tybjærg-Hansen, A, Haverkate, F, Lee, AJ, Smith, FB, Salomaa, V, Harald, K, Rasi, V, Jousilahti, P, Pekkanen, J, D'Agostino, R, Wilson, PWF, Tofler, G, Levy, D, Marchioli, R, Valagussa, F, Rosengren, A, Lappas, G, Eriksson, H, Cremer, P, Nagel, D, Curb, JD, Rodriguez, B, Yano, K, Salonen, JT, Nyyssönen, K, Tuomainen, TP, Hedblad, B, Engström, G, Berglund, G, Loewel, H, Hense, HW, Mead, TW, Cooper, JA, De Stavola, B, Knottenbelt, C, Miller, GJ, Cooper, JA, Bauer, KA, Rosenberg, RD, Sato, S, Kitamura, A, Naito, Y, Iso, H, Harald, K, Rasi, V, Vahtera, E, Jousilahti, P, Palosuo, T, Ducimetriere, P, Amouyel, P, Arveiler, D, Evans, AE, Ferrieres, J, Juhan-Vague, I, Bingham, A, Schulte, H, Assmann, G, Cantin, B, Lamarche, B, Després, JP, Tunstall-Pedoe, H, Lowe, GDO, Woodward, M, Ben-Shlomo, Y, Smith, GD, Palmieri, V, Yeh, JL, Meade, TW, Brennan, P, Knottenbelt, C, Ridker, P, Rodeghiero, F, Tosetto, A, Shepherd, J, Lowe, GDO, Ford, I, Robertson, M, Brunner, E, Shipley, M, Feskens, EJM, Kromhout, D, Di Angelantonio, E, Kaptoge, S, Lewington, S & Lowe, GDO 2009, 'Correcting for multivariate measurement error by regression calibration in meta-analyses of epidemiological studies', Statistics in Medicine, vol. 28, no. 7, pp. 1067-1092. https://doi.org/10.1002/sim.3530

TY - JOUR

T1 - Correcting for multivariate measurement error by regression calibration in meta-analyses of epidemiological studies

AU - Wood, Angela Mary

AU - Thompson, S. G.

AU - Kostis, J. B.

AU - Wilson, A. C.

AU - Wu, K.

AU - Benderly, M.

AU - Goldbourt, U.

AU - Willeit, J.

AU - Kiechl, S.

AU - Yarnell, J. W.G.

AU - Sweetman, P. M.

AU - Elwood, P. C.

AU - Cushman, M.

AU - Tracy, R. P.

AU - Tybjærg-Hansen, A.

AU - Haverkate, F.

AU - Lee, A. J.

AU - Smith, F. B.

AU - Salomaa, V.

AU - Harald, K.

AU - Rasi, V.

AU - Jousilahti, P.

AU - Pekkanen, J.

AU - D'Agostino, R.

AU - Wilson, P. W.F.

AU - Tofler, G.

AU - Levy, D.

AU - Marchioli, R.

AU - Valagussa, F.

AU - Rosengren, A.

AU - Lappas, G.

AU - Eriksson, H.

AU - Cremer, P.

AU - Nagel, D.

AU - Curb, J. D.

AU - Rodriguez, B.

AU - Yano, K.

AU - Salonen, J. T.

AU - Nyyssönen, K.

AU - Tuomainen, T. P.

AU - Hedblad, B.

AU - Engström, G.

AU - Berglund, G.

AU - Loewel, H.

AU - Hense, H. W.

AU - Mead, T. W.

AU - Cooper, J. A.

AU - De Stavola, B.

AU - Knottenbelt, C.

AU - Miller, G. J.

AU - Cooper, J. A.

AU - Bauer, K. A.

AU - Rosenberg, R. D.

AU - Sato, S.

AU - Kitamura, A.

AU - Naito, Y.

AU - Iso, H.

AU - Harald, K.

AU - Rasi, V.

AU - Vahtera, E.

AU - Jousilahti, P.

AU - Palosuo, T.

AU - Ducimetriere, P.

AU - Amouyel, P.

AU - Arveiler, D.

AU - Evans, A. E.

AU - Ferrieres, J.

AU - Juhan-Vague, I.

AU - Bingham, A.

AU - Schulte, H.

AU - Assmann, G.

AU - Cantin, B.

AU - Lamarche, B.

AU - Després, J. P.

AU - Tunstall-Pedoe, H.

AU - Lowe, G. D.O.

AU - Woodward, M.

AU - Ben-Shlomo, Y.

AU - Smith, G. Davey

AU - Palmieri, V.

AU - Yeh, J. L.

AU - Meade, T. W.

AU - Brennan, P.

AU - Knottenbelt, C.

AU - Ridker, P.

AU - Rodeghiero, F.

AU - Tosetto, A.

AU - Shepherd, J.

AU - Lowe, G. D.O.

AU - Ford, I.

AU - Robertson, M.

AU - Brunner, E.

AU - Shipley, M.

AU - Feskens, E. J.M.

AU - Kromhout, D.

AU - Di Angelantonio, E.

AU - Kaptoge, S.

AU - Lewington, S.

AU - Lowe, G. D.O.

PY - 2009/3/30

Y1 - 2009/3/30

N2 - Within-person variability in measured values of multiple risk factors can bias their associations with disease. The multivariate regression calibration (RC) approach can correct for such measurement error and has been applied to studies in which true values or independent repeat measurements of the risk factors are observed on a subsample. We extend the multivariate RC techniques to a meta-analysis framework where multiple studies provide independent repeat measurements and information on disease outcome. We consider the cases where some or all studies have repeat measurements, and compare study-specific, averaged and empirical Bayes estimates of RC parameters. Additionally, we allow for binary covariates (e.g. smoking status) and for uncertainty and time trends in the measurement error corrections. Our methods are illustrated using a subset of individual participant data from prospective long-term studies in the Fibrinogen Studies Collaboration to assess the relationship between usual levels of plasma fibrinogen and the risk of coronary heart disease, allowing for measurement error in plasma fibrinogen and several confounders.

AB - Within-person variability in measured values of multiple risk factors can bias their associations with disease. The multivariate regression calibration (RC) approach can correct for such measurement error and has been applied to studies in which true values or independent repeat measurements of the risk factors are observed on a subsample. We extend the multivariate RC techniques to a meta-analysis framework where multiple studies provide independent repeat measurements and information on disease outcome. We consider the cases where some or all studies have repeat measurements, and compare study-specific, averaged and empirical Bayes estimates of RC parameters. Additionally, we allow for binary covariates (e.g. smoking status) and for uncertainty and time trends in the measurement error corrections. Our methods are illustrated using a subset of individual participant data from prospective long-term studies in the Fibrinogen Studies Collaboration to assess the relationship between usual levels of plasma fibrinogen and the risk of coronary heart disease, allowing for measurement error in plasma fibrinogen and several confounders.

KW - Individual participant data

KW - Measurement error

KW - Meta-analysis

KW - Regression calibration

KW - Within-person variation

UR - http://www.scopus.com/inward/record.url?scp=67749102677&partnerID=8YFLogxK

U2 - 10.1002/sim.3530

DO - 10.1002/sim.3530

M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???

AN - SCOPUS:67749102677

SN - 0277-6715

VL - 28

SP - 1067

EP - 1092

JO - Statistics in Medicine

JF - Statistics in Medicine

IS - 7

ER -

Correcting for multivariate measurement error by regression calibration in meta-analyses of epidemiological studies

Abstract

Funding

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this