Project

Back to overview

Predicting CVD mortality: Can self-reported parameters compensate for lacking information on classical risk factors?

English title Predicting CVD mortality: Can self-reported parameters compensate for lacking information on classical risk factors?
Applicant Gutzwiller Felix
Number 143897
Funding scheme Project funding (special)
Research institution Institut für Epidemiologie, Biostatistik und Prävention Universität Zürich
Institution of higher education University of Zurich - ZH
Main discipline Cardiovascular Diseases
Start/End 01.10.2012 - 31.07.2013
Approved amount 121'504.00
Show all

All Disciplines (5)

Discipline
Cardiovascular Diseases
Cardiovascular Research
Medical Statistics
Public Health and Health Services
Methods of Epidemiology and Preventive Medicine

Keywords (13)

Cardiovascular disease (CVD); CVD risk scores; Coronary Heart Disease (CHD); stroke; risk factors; self-perceived health; Mortality; Swiss National Cohort; large nested project; Calibration; self-reported data; survey; Participation bias

Lay Summary (German)

Lead
Klassische klassische Herz-Kreislauf-Risikofaktoren setzen eine Blutentnahme mit Labormessung voraus. Viele Menschen kommen nicht in den Genuss einer Risikoeinschätzung, weil diese Werte nie erhoben wurden. Der innovative Ansatz einer Risikoermittlung mittels Selbstangaben ist ein vielversprechender Ansatz um dieser Einschränkung entgegenzuwirken.
Lay summary

Einschätzung des Herz-Kreislauferkrankungsrisikos: Auch möglich mit Selbstangaben, wenn Angaben zu klinischen Risikofaktoren fehlen?

Traditionell wird das globale Risiko einer Person, einen Herzinfarkt oder einen Hirnschlag zu erleiden, mit Hilfe von sogenannt klinischen Herz-Kreislauf-Risikofaktoren ermittelt. Dazu gehören der Blutdruck und die Cholesterinwerte. Diese Angaben setzen aber eine Untersuchung mit einem Messgerät und eine Blutentnahme mit Labormessung voraus. Der Nachteil dieser Methode ist, dass viele Menschen nicht in den Genuss einer Risikoeinschätzung kommen, weil diese Werte bei ihnen nicht bekannt sind. Damit geht Präventionspotential verloren, weil eine solche Standortbestimmung dazu motivieren kann, den eigenen Lebensstil zu verbessern. 

Diese Studie untersucht, wie gut Faktoren, die nur mittels Befragung erfasst wurden, im Vergleich zu traditionellen, klinisch erhobenen Risikofaktoren abschneiden. Ziel ist es Faktoren zu ermitteln und zu testen, die, in Kombination angewendet, eine ähnlich gute Voraussage eines Herz-Kreislauf-Ereignisses erlauben wie die traditionelle Risikofaktoren. Kandidaten für solche Faktoren sind beispielsweise selbst angegebene Grösse und Gewicht, die die Berechnung des Body-Mass-Index ermöglichen.

Grundlage für diese Studie bietet die Swiss National Cohort, die Sterbedaten mit Daten der Volkszählungen verbindet und damit Informationen zu soziodemografischen Einflussfaktoren liefert. Diese Daten werden anonym über demografische Variablen (Geschlecht, Geburtsdatum, Wohnort etc.) mit den Daten aus fünf Studien aus den 1970er, 1980er und 1990er Jahren verknüpft, die eine Vielzahl von gesundheitsrelevanten Mess- und Befragungsdaten enthalten. Die lange Beobachtungszeit und die grosse Zahl an Teilnehmern und Sterbefällen erlaubt nicht nur Berechnungen der Überlebenszeit und der Sterberisiken unter Schweizer Gegebenheiten, sondern auch Evaluation und Anpassung von bestehenden Herz-Kreislauf-Risikoschätzern.

In der Schweiz fehlt für solche Analysen eine national repräsentative prospektive Studie aus der Allgemeinbevölkerung. Diese Studie bietet einen kosteneffizienten Teilersatz dafür und erlaubt damit neue Erkenntnisse über den Zusammenhang zwischen Herz-Kreislauf-Risiken und der Sterblichkeit in der Schweiz. Der innovative Ansatz von Risikoermittlung mittels Selbstangaben dürfte auch international Beachtung finden. Die erwarteten Resultate dürften nicht nur für die Allgemeinbevölkerung von Bedeutung sein, sondern auch ÄrztInnen eine nützliches Werkzeug bieten für die Beurteilung ihrer PatientInnen bzw. für deren Motivation zu einer Verbesserung ihres Lebensstils.


Direct link to Lay Summary Last update: 03.09.2013

Responsible applicant and co-applicants

Employees

Publications

Publication
Better risk assessment with glycated hemoglobin instead of cholesterol in CVD risk prediction charts
Faeh D, Rohrmann S, Braun J (2013), Better risk assessment with glycated hemoglobin instead of cholesterol in CVD risk prediction charts, in European Journal of Epidemiology, 1-5.
Blood glucose may be an alternative to cholesterol in CVD risk prediction charts
Braun J, Bopp M, Faeh D (2013), Blood glucose may be an alternative to cholesterol in CVD risk prediction charts, in Cardiovascular Diabetology, 12(1), 0-0.
Direct estimation of death attributable to smoking in Switzerland based on record linkage of routine and observational data
Maag J, Braun J, Bopp M, Faeh D (2013), Direct estimation of death attributable to smoking in Switzerland based on record linkage of routine and observational data, in Nicotine and Tobacco Research, 15(9), 1588-1597.
Population specific and up to date cardiovascular risk charts can be efficiently obtained with record linkage of routine and observational data
Faeh D, Braun J, Rufibach K, Puhan M, Marques-Vidal P, Bopp M (2013), Population specific and up to date cardiovascular risk charts can be efficiently obtained with record linkage of routine and observational data, in PloS One, 8(2), 0-0.
Body mass index vs cholesterol in cardiovascular disease risk prediction models
Faeh D, Braun J, Bopp M (2012), Body mass index vs cholesterol in cardiovascular disease risk prediction models, in Archives of Internal Medicine, 172(22), 1766-1768.
Inverse association between circulating vitamin D and mortality – dependent on sex and cause of death?
Rohrmann S, Braun J, Bopp M, Faeh D, Inverse association between circulating vitamin D and mortality – dependent on sex and cause of death?, in Nutrition Metabolism and Cardiovascular Disease.
J-shaped association between plasma glucose concentration and cardiovascular disease mortality over a follow-up of 32 years
Von Gunten E, Braun J, Bopp M, Faeh D, J-shaped association between plasma glucose concentration and cardiovascular disease mortality over a follow-up of 32 years, in Preventive Medicine.

Collaboration

Group / person Country
Types of collaboration
John Hopkins Bloomberg School of Public Health United States of America (North America)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
University of Lausanne Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
In the framework of the Swiss National Cohort Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
AACR (Am Assoc Cancer Research) 06.04.2013 Washington DC
DGE (Dt. Ges. Ernährung) 20.03.2013 Bonn


Associated projects

Number Title Start Funding scheme
134273 The Swiss National Cohort: a platform for longitudinal research in Switzerland 01.07.2011 Cohort Studies Large
125710 Impact of risk factors on cardiovascular mortality: Results of 30 years follow-up 01.06.2009 Project funding (special)

Abstract

Background.More than three decades ago, Professor Felix Gutzwiller started his research career by implementing the first National Research Program (NRP 1A), aimed at the primary prevention of cardiovascular disease (CVD). More pioneering work followed, e.g. the first national health survey or the Swiss National Cohort (SNC), both of which were funded by the SNSF. Nevertheless, and despite all effort and progresses, no population-based cohort, allowing examining the impact of CVD risk factors on health in Switzerland, could be established. With the help of the SNSF (grant 32473B_125710, June 2009 - May 2012), we were able to partially compensate this research lacuna. Within this time frame, we were able to conduct the core tasks described in the proposal. Two large population based historical studies conducted between 1977 and 1993 with self-administered questionnaire and medical exam data could be linked with the SNC. Including deaths until 2008 allowed to follow-up 18,000 participants for up to 30 years. Various publications proved the potential of this database to provide original and valuable findings. However, due to technical and administrative reasons, we were not able to achieve all objectives described in our proposal within runtime of grant 32473B_125710. First of all this concerns a substantial delay in the record linkage of SNC with two national health surveys, which will not be finished before April/May 2012. However, in these three years, several new promising research questions arose, which call for being tested with this unique database.Specific aims. a) Enhancing and complementing research questions:•To evaluate whether CVD mortality can be predicted by self-reported risk factors and to derive multivariate risk models and to test how this prediction performs compared to classical risk algorithms•To examine whether the predictive power of blood pressure on CVD mortality can be improved by including other circulatory health parameters such as heart rate or heart rate variability•To test whether a single blood glucose measurement is independently associated with mortality and, if yes, how long this association persistsb) Pursuing new research questions:•To explore whether the health promoting effect of the Mediterranean diet varies by cultural/migrational background•To estimate the impact of different domains of physical activity on CVD mortality •To compare CVD mortality among persons who agreed and declined to participate in health studies for different reasons and to analyse whether these patterns depend on socio-demographic parameters and kind of recruitmentMethods. Data on diet and physical activity will be selected and categorized in order to allow comparability with other research. Potential candidates for the establishment of risk scores based on measured and/or self-reported risk factors will be evaluated and tested. Poisson, logistic, Cox and Weibull regression will be used to examine the independent effects of risk factors on mortality. Based on the results of a Weibull regression model, a risk score will be calculated allowing the estimation of death from CVD within 10 and 20 years associated with a specific CVD risk factors constellation. Other possible variables for the generation of risk scores will be evaluated and tested. For discrimination and calibration of risk scores, C statistics will be applied as well as the Brier score. Novel metrics for model performance such as net reclassification tables or decision curve analysis will also be considered. Significance. The linkage success of the historical studies with the SNC has exceeded the expectations. However, the procedure was cumbersome and time consuming. Although several papers could successfully be published in renowned international peer-reviewed journals, the potential of the generated data by far could not be exploited. With its support, the SNSF could further valorise the past investment in the involved health studies and the setup and maintenance of the SNC and thus increase the return of investment in these projects. Finally, the completion of the proposed project would enable to put the capstone on the efforts of Professor Felix Gutzwiller to found and establish in Switzerland an appropriate environment for internationally competitive research in CVD epidemiology and public health, 35 years after being appointed as project director of the NRP 1A.
-