Copyright © 2003 by the European Society of Cardiology.
Letter to the Editor
Response to the letter by Müller-Nordhorn and Willich
Institute of Epidemiology & Social Medicine, University of Muenster, Muenster, Germany
* Correspondence to: Tel: +49-251-8355399; Fax: +49 251 835 5300
E-mail address: hense{at}uni-muenster.de
Dear Sir,
We thank Drs Müller-Nordhorn and Willich for their interest in our paper. They critically question the validity of our conclusion that the Framingham risk function overestimates risk of coronary heart disease in men and women from Germany. Firstly, they assume a potentially strong healthy worker effect among the participants of the PROCAM study. It is worth noting that the estimate they present for this effect—25% lower cardiovascular risk-was derived in a group of industrial workers where selective forces in favour of the physically healthy, at hire and by time on employment, is known to be particularly pronounced. However, PROCAM was not an industrial worksite screening: the PROCAM sample is composed of almost 15% civil servants, more than 50% employees and only about one-third of workers. Further, PROCAM risk estimates, if anything, were higher than those from the general population sample of MONICA/KORA (table 3). Next, Müller-Nordhorn and Willich point to the fact that regional variation of risk is a problem. This is true-for any country with a sizable population and geographic expansion (see e.g. reference 30 in our paper which relates to ethnic groups in the US). However, we cannot follow the argument that data from recent German prospective population studies are of no valid use for assessing absolute CHD risk while at the same time validity is claimed for a far less recent study in subjects sampled from a single town situated on the east coast of the United States of America. Müller-Nordhorn and Willich criticize further our interpretation of table 5 as ‘confirming a fairly consistent ratio of 2–3 across the age groups’. In fact, confidence intervals were omitted from this table for better legibility and comprehensibility: the wide confidence intervals, in particular among the youngest women, overlapped widely and did not give any indication of significant differences between the ratios. We concede, however, that our presentation may have been misleading. Another critique concerns the use of risk prediction equations derived from subjects aged 30 to 74 years in the Framingham Heart Study to the restricted age range of our samples. We seem to be unable to follow the reasoning in this point because Framingham risk predictions were devised to be used in exactly the age groups that we investigated, that is, 35 to 64 years of age. Speculations about CHD risk in the German elderly and its prediction go beyond the evidence available in this study. Finally, we alluded to the PROCAM score only to indicate that there areoptions worthy of consideration.
In conclusion, we emphasize our point that data which originate from recent prospective studies executed within Germany are appropriate to validly assess the level of absolute CHD risk in this country. We further maintain our conclusion that Framingham risk equations overestimate risk of German populations and exemplify this with results from two recent prospective studies. In our view there is a need of correcting Framingham-based estimates of absolute risk by use of locally adapted risk charts to avoid ‘risk inflation’ in primary prevention.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||