Why and how to control for age in occupational epidemiology

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

In occupational epidemiology, the need to consider the age factor properly influences the choice of study design and analytical techniques. In most studies, age is viewed as a potential confounder. Age is strongly associated with end points of interest in occupational epidemiology (diseases, physiological characteristics, doses of xenobiotics, etc), but to measure age as a confounder it must be associated with the exposure under study. When the exposure of interest is time related - for example, duration of employment, time since first exposure, cumulative exposure - a strong intrinsic association with age can be anticipated, and age will behave as a (usually strong) confounder. When occupational exposures without a direct relation with age - for example, job, department, type of exposure-are evaluated, the degree and direction of confounding bias cannot be anticipated. Control of the confounding effect of age can be accomplished in the design phase of a study by way of randomisation, restriction, and matching. Randomisation is seldom viable in occupational settings. Restriction is rarely used in the case of age. Matching is often used in a case-control study as a method to increase the study efficiency, but it must be followed by proper matched or stratified analysis. Options for age adjustment in the analysis phase involve stratification and regression methods. In longitudinal studies the modified life table analysis is used to take into account the fact that subjects cross categories of age as the study proceeds. Stability of relative measures of effect over age strata favoured the greater use of relative risks than risk differences. In the presence of effect modification the influence of age should not be eliminated; its interaction with exposure should be explicitly considered.

Original languageEnglish
Pages (from-to)772-776
Number of pages5
JournalOccupational and Environmental Medicine
Volume54
Issue number11
Publication statusPublished - 1997

Fingerprint

epidemiology
Random Allocation
Epidemiology
Life Tables
Occupational Diseases
Age Factors
Xenobiotics
Occupational Exposure
Longitudinal Studies
Case-Control Studies
life table
occupational exposure
xenobiotics
Direction compound
exposure
analytical method
stratification

Keywords

  • Age
  • Confounding bias
  • Occupational epidemiology

ASJC Scopus subject areas

  • Public Health, Environmental and Occupational Health
  • Environmental Science(all)

Cite this

Why and how to control for age in occupational epidemiology. / Consonni, Dario.

In: Occupational and Environmental Medicine, Vol. 54, No. 11, 1997, p. 772-776.

Research output: Contribution to journalArticle

@article{58e481edfc304627b826ed64b6dc59a3,
title = "Why and how to control for age in occupational epidemiology",
abstract = "In occupational epidemiology, the need to consider the age factor properly influences the choice of study design and analytical techniques. In most studies, age is viewed as a potential confounder. Age is strongly associated with end points of interest in occupational epidemiology (diseases, physiological characteristics, doses of xenobiotics, etc), but to measure age as a confounder it must be associated with the exposure under study. When the exposure of interest is time related - for example, duration of employment, time since first exposure, cumulative exposure - a strong intrinsic association with age can be anticipated, and age will behave as a (usually strong) confounder. When occupational exposures without a direct relation with age - for example, job, department, type of exposure-are evaluated, the degree and direction of confounding bias cannot be anticipated. Control of the confounding effect of age can be accomplished in the design phase of a study by way of randomisation, restriction, and matching. Randomisation is seldom viable in occupational settings. Restriction is rarely used in the case of age. Matching is often used in a case-control study as a method to increase the study efficiency, but it must be followed by proper matched or stratified analysis. Options for age adjustment in the analysis phase involve stratification and regression methods. In longitudinal studies the modified life table analysis is used to take into account the fact that subjects cross categories of age as the study proceeds. Stability of relative measures of effect over age strata favoured the greater use of relative risks than risk differences. In the presence of effect modification the influence of age should not be eliminated; its interaction with exposure should be explicitly considered.",
keywords = "Age, Confounding bias, Occupational epidemiology",
author = "Dario Consonni",
year = "1997",
language = "English",
volume = "54",
pages = "772--776",
journal = "Occupational and Environmental Medicine",
issn = "1351-0711",
publisher = "BMJ Publishing Group",
number = "11",

}

TY - JOUR

T1 - Why and how to control for age in occupational epidemiology

AU - Consonni, Dario

PY - 1997

Y1 - 1997

N2 - In occupational epidemiology, the need to consider the age factor properly influences the choice of study design and analytical techniques. In most studies, age is viewed as a potential confounder. Age is strongly associated with end points of interest in occupational epidemiology (diseases, physiological characteristics, doses of xenobiotics, etc), but to measure age as a confounder it must be associated with the exposure under study. When the exposure of interest is time related - for example, duration of employment, time since first exposure, cumulative exposure - a strong intrinsic association with age can be anticipated, and age will behave as a (usually strong) confounder. When occupational exposures without a direct relation with age - for example, job, department, type of exposure-are evaluated, the degree and direction of confounding bias cannot be anticipated. Control of the confounding effect of age can be accomplished in the design phase of a study by way of randomisation, restriction, and matching. Randomisation is seldom viable in occupational settings. Restriction is rarely used in the case of age. Matching is often used in a case-control study as a method to increase the study efficiency, but it must be followed by proper matched or stratified analysis. Options for age adjustment in the analysis phase involve stratification and regression methods. In longitudinal studies the modified life table analysis is used to take into account the fact that subjects cross categories of age as the study proceeds. Stability of relative measures of effect over age strata favoured the greater use of relative risks than risk differences. In the presence of effect modification the influence of age should not be eliminated; its interaction with exposure should be explicitly considered.

AB - In occupational epidemiology, the need to consider the age factor properly influences the choice of study design and analytical techniques. In most studies, age is viewed as a potential confounder. Age is strongly associated with end points of interest in occupational epidemiology (diseases, physiological characteristics, doses of xenobiotics, etc), but to measure age as a confounder it must be associated with the exposure under study. When the exposure of interest is time related - for example, duration of employment, time since first exposure, cumulative exposure - a strong intrinsic association with age can be anticipated, and age will behave as a (usually strong) confounder. When occupational exposures without a direct relation with age - for example, job, department, type of exposure-are evaluated, the degree and direction of confounding bias cannot be anticipated. Control of the confounding effect of age can be accomplished in the design phase of a study by way of randomisation, restriction, and matching. Randomisation is seldom viable in occupational settings. Restriction is rarely used in the case of age. Matching is often used in a case-control study as a method to increase the study efficiency, but it must be followed by proper matched or stratified analysis. Options for age adjustment in the analysis phase involve stratification and regression methods. In longitudinal studies the modified life table analysis is used to take into account the fact that subjects cross categories of age as the study proceeds. Stability of relative measures of effect over age strata favoured the greater use of relative risks than risk differences. In the presence of effect modification the influence of age should not be eliminated; its interaction with exposure should be explicitly considered.

KW - Age

KW - Confounding bias

KW - Occupational epidemiology

UR - http://www.scopus.com/inward/record.url?scp=0030696236&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030696236&partnerID=8YFLogxK

M3 - Article

VL - 54

SP - 772

EP - 776

JO - Occupational and Environmental Medicine

JF - Occupational and Environmental Medicine

SN - 1351-0711

IS - 11

ER -