Thromb Haemost 2005; 93(05): 955-963
DOI: 10.1160/TH04-12-0805
Cell Signalling and Vessel Remodelling
Schattauer GmbH

Associations of fibrinogen and C-reactive protein with prevalent and incident coronary heart disease are attenuated by adjustment for confounding factors

British Women’s Heart and Health Study
Debbie A. Lawlor
1   Department of Social Medicine, University of Bristol, Bristol, UK
,
George Davey Smith
1   Department of Social Medicine, University of Bristol, Bristol, UK
,
Ann Rumley
2   Division of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, UK
,
Gordon D. O. Lowe
2   Division of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, UK
,
Shah Ebrahim
1   Department of Social Medicine, University of Bristol, Bristol, UK
› Author Affiliations
Finiancial support: We thank the British Heart Foundation for financial support (Grant PG/02/136) and the (UK) Department of Health for core support to the British Women’s Heart & Health Study. DAL is funded by a (UK) Department of Health Career Scientist Award.
Further Information

Publication History

Received 15 December 2004

Accepted after resubmission 14 February 2005

Publication Date:
11 December 2017 (online)

Zoom Image

Summary

A cross sectional and prospective analysis of 3,745 British women aged 60–79 years at baseline was undertaken. Among these women there were 570 prevalent cases of coronary heart disease (CHD) and 151 new cases among 12,641 person-years of follow up of women who were free of CHD at baseline. Both fibrinogen and CRP were associated with indicators of socioeconomic position in childhood and adulthood and there was a cumulative effect of socioeconomic position from across the life course. The age-adjusted odds ratio (95% confidence interval) of prevalent CHD for a 1 unit (1 g/L) increase in fibrinogen was 1.29 (1.12, 1.49); with full adjustment for all potential confounding factors this attenuated to 1.09 (0.93, 1.28). The hazards ratio for incident CHD among those free of disease at baseline was 1.28 (1.00, 1.64); with full adjustment for all potential confounding factors this attenuated to 1.09 (0.84, 1.44). Similar effects of adjustment for confounding factors were seen for the associations between CRP and both prevalent and incident CHD. By contrast, the strong positive association between smoking (an established causal risk factor for CHD) and CHD was not attenuated by adjustment for life course socioeconomic position or other risk factors. We conclude that fibrinogen and CRP predict CHD but may not be causally related to it.