Estimating intervention effects in longitudinal studies

James H. Dwyer, David Mackinnon, Mary Ann Pentz, Brian R. Flay, William B. Hansen, Eric Yu I Wang, C. Anderson Johnson

Research output: Contribution to journalArticle

31 Citations (Scopus)

Abstract

Longitudinal studies aimed at assessing the impact of interventions on disease risk factors often confront several statistical problems. These problems include 1) dependent variables measured by ordered categories, 2) numerous potentially relevant patterns of transition between outcome levels, 3) mixed units of analysis (e.g., assignment by social unit while theorizing in terms of individuals), 4) incomplete randomization, and 5) correlated estimates for successive occasions of longitudinal measurement Longitudinal data on use of cigarettes, alcohol, and marijuana among adolescents (n = 1,244, complete data) from the Midwestern Prevention Project are used to demonstrate solutions to each of these problems: 1) a proportional odds regression model, 2) conditional logistic models of transitions with interactions between baseline level and intervention effect, 3) a logistic model estimated with linear regression methods on measures aggregated by social unit, 4) conditional and unconditional models of effect magnitude, and 5) a repeated measures logistic regression technique. Panel data fit to the various models yielded the following conclusions concerning intervention effects in the Midwestern Prevention Project reduction in the prevalence of cigarette users in treatment schools compared with control schools (8% vs. 18% smoked in the last week at one year follow-up), mixed evidence of an effect on marijuana use, and no evidence of an effect on alcohol use.

Original languageEnglish (US)
Pages (from-to)781-796
Number of pages16
JournalAmerican Journal of Epidemiology
Volume130
Issue number4
StatePublished - Nov 1989
Externally publishedYes

Fingerprint

Longitudinal Studies
Logistic Models
Cannabis
Tobacco Products
Alcohols
Random Allocation
Linear Models
Therapeutics

Keywords

  • Longitudinal studies
  • Primary prevention
  • Regression analysis
  • Research design
  • Smoking
  • Substance abuse

ASJC Scopus subject areas

  • Geriatrics and Gerontology
  • Epidemiology

Cite this

Dwyer, J. H., Mackinnon, D., Pentz, M. A., Flay, B. R., Hansen, W. B., Wang, E. Y. I., & Johnson, C. A. (1989). Estimating intervention effects in longitudinal studies. American Journal of Epidemiology, 130(4), 781-796.

Estimating intervention effects in longitudinal studies. / Dwyer, James H.; Mackinnon, David; Pentz, Mary Ann; Flay, Brian R.; Hansen, William B.; Wang, Eric Yu I; Johnson, C. Anderson.

In: American Journal of Epidemiology, Vol. 130, No. 4, 11.1989, p. 781-796.

Research output: Contribution to journalArticle

Dwyer, JH, Mackinnon, D, Pentz, MA, Flay, BR, Hansen, WB, Wang, EYI & Johnson, CA 1989, 'Estimating intervention effects in longitudinal studies', American Journal of Epidemiology, vol. 130, no. 4, pp. 781-796.
Dwyer JH, Mackinnon D, Pentz MA, Flay BR, Hansen WB, Wang EYI et al. Estimating intervention effects in longitudinal studies. American Journal of Epidemiology. 1989 Nov;130(4):781-796.
Dwyer, James H. ; Mackinnon, David ; Pentz, Mary Ann ; Flay, Brian R. ; Hansen, William B. ; Wang, Eric Yu I ; Johnson, C. Anderson. / Estimating intervention effects in longitudinal studies. In: American Journal of Epidemiology. 1989 ; Vol. 130, No. 4. pp. 781-796.
@article{2d74271fdc0a46228d65b4cc2391a4ad,
title = "Estimating intervention effects in longitudinal studies",
abstract = "Longitudinal studies aimed at assessing the impact of interventions on disease risk factors often confront several statistical problems. These problems include 1) dependent variables measured by ordered categories, 2) numerous potentially relevant patterns of transition between outcome levels, 3) mixed units of analysis (e.g., assignment by social unit while theorizing in terms of individuals), 4) incomplete randomization, and 5) correlated estimates for successive occasions of longitudinal measurement Longitudinal data on use of cigarettes, alcohol, and marijuana among adolescents (n = 1,244, complete data) from the Midwestern Prevention Project are used to demonstrate solutions to each of these problems: 1) a proportional odds regression model, 2) conditional logistic models of transitions with interactions between baseline level and intervention effect, 3) a logistic model estimated with linear regression methods on measures aggregated by social unit, 4) conditional and unconditional models of effect magnitude, and 5) a repeated measures logistic regression technique. Panel data fit to the various models yielded the following conclusions concerning intervention effects in the Midwestern Prevention Project reduction in the prevalence of cigarette users in treatment schools compared with control schools (8{\%} vs. 18{\%} smoked in the last week at one year follow-up), mixed evidence of an effect on marijuana use, and no evidence of an effect on alcohol use.",
keywords = "Longitudinal studies, Primary prevention, Regression analysis, Research design, Smoking, Substance abuse",
author = "Dwyer, {James H.} and David Mackinnon and Pentz, {Mary Ann} and Flay, {Brian R.} and Hansen, {William B.} and Wang, {Eric Yu I} and Johnson, {C. Anderson}",
year = "1989",
month = "11",
language = "English (US)",
volume = "130",
pages = "781--796",
journal = "American Journal of Epidemiology",
issn = "0002-9262",
publisher = "Oxford University Press",
number = "4",

}

TY - JOUR

T1 - Estimating intervention effects in longitudinal studies

AU - Dwyer, James H.

AU - Mackinnon, David

AU - Pentz, Mary Ann

AU - Flay, Brian R.

AU - Hansen, William B.

AU - Wang, Eric Yu I

AU - Johnson, C. Anderson

PY - 1989/11

Y1 - 1989/11

N2 - Longitudinal studies aimed at assessing the impact of interventions on disease risk factors often confront several statistical problems. These problems include 1) dependent variables measured by ordered categories, 2) numerous potentially relevant patterns of transition between outcome levels, 3) mixed units of analysis (e.g., assignment by social unit while theorizing in terms of individuals), 4) incomplete randomization, and 5) correlated estimates for successive occasions of longitudinal measurement Longitudinal data on use of cigarettes, alcohol, and marijuana among adolescents (n = 1,244, complete data) from the Midwestern Prevention Project are used to demonstrate solutions to each of these problems: 1) a proportional odds regression model, 2) conditional logistic models of transitions with interactions between baseline level and intervention effect, 3) a logistic model estimated with linear regression methods on measures aggregated by social unit, 4) conditional and unconditional models of effect magnitude, and 5) a repeated measures logistic regression technique. Panel data fit to the various models yielded the following conclusions concerning intervention effects in the Midwestern Prevention Project reduction in the prevalence of cigarette users in treatment schools compared with control schools (8% vs. 18% smoked in the last week at one year follow-up), mixed evidence of an effect on marijuana use, and no evidence of an effect on alcohol use.

AB - Longitudinal studies aimed at assessing the impact of interventions on disease risk factors often confront several statistical problems. These problems include 1) dependent variables measured by ordered categories, 2) numerous potentially relevant patterns of transition between outcome levels, 3) mixed units of analysis (e.g., assignment by social unit while theorizing in terms of individuals), 4) incomplete randomization, and 5) correlated estimates for successive occasions of longitudinal measurement Longitudinal data on use of cigarettes, alcohol, and marijuana among adolescents (n = 1,244, complete data) from the Midwestern Prevention Project are used to demonstrate solutions to each of these problems: 1) a proportional odds regression model, 2) conditional logistic models of transitions with interactions between baseline level and intervention effect, 3) a logistic model estimated with linear regression methods on measures aggregated by social unit, 4) conditional and unconditional models of effect magnitude, and 5) a repeated measures logistic regression technique. Panel data fit to the various models yielded the following conclusions concerning intervention effects in the Midwestern Prevention Project reduction in the prevalence of cigarette users in treatment schools compared with control schools (8% vs. 18% smoked in the last week at one year follow-up), mixed evidence of an effect on marijuana use, and no evidence of an effect on alcohol use.

KW - Longitudinal studies

KW - Primary prevention

KW - Regression analysis

KW - Research design

KW - Smoking

KW - Substance abuse

UR - http://www.scopus.com/inward/record.url?scp=0024451122&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0024451122&partnerID=8YFLogxK

M3 - Article

C2 - 2773920

AN - SCOPUS:0024451122

VL - 130

SP - 781

EP - 796

JO - American Journal of Epidemiology

JF - American Journal of Epidemiology

SN - 0002-9262

IS - 4

ER -