• Open Access

Is the use of cholesterol in mortality risk algorithms in clinical guidelines valid? Ten years prospective data from the Norwegian HUNT 2 study

Authors

  • Halfdan Petursson MD,

    Corresponding author
    1. Research Fellow, Research Unit of General Practice, Department of Public Health and General Practice, Norwegian University of Science and Technology (NTNU), Trondheim, Norway
      Dr Halfdan Petursson, Research Unit of General Practice, Department of Public Health and General Practice, Norwegian University of Science and Technology (NTNU), PO Box 8905, 7491 Trondheim, Norway, E-mail: halfdanpe@gmail.com
    Search for more papers by this author
  • Johann A. Sigurdsson MD Dr med,

    1. Professor, Department of Family Medicine, University of Iceland, and Centre of Development, Primary Health Care of the Capital Area, Reykjavik, Iceland
    Search for more papers by this author
  • Calle Bengtsson MD Dr med,

    1. Professor Emeritus, Department of Primary Health Care, The Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
    Search for more papers by this author
  • Tom I. L. Nilsen Dr Philos,

    1. Associate Professor, Department of Human Movement Science, Norwegian University of Science and Technology (NTNU), Trondheim, Norway
    Search for more papers by this author
  • Linn Getz MD PhD

    1. Associate Professor, Research Unit of General Practice, Department of Public Health and General Practice, Norwegian University of Science and Technology (NTNU), Trondheim, Norway and Landspitali University Hospital, Reykjavik, Iceland
    Search for more papers by this author

Dr Halfdan Petursson, Research Unit of General Practice, Department of Public Health and General Practice, Norwegian University of Science and Technology (NTNU), PO Box 8905, 7491 Trondheim, Norway, E-mail: halfdanpe@gmail.com

Abstract

Rationale, aims and objectives  Many clinical guidelines for cardiovascular disease (CVD) prevention contain risk estimation charts/calculators. These have shown a tendency to overestimate risk, which indicates that there might be theoretical flaws in the algorithms. Total cholesterol is a frequently used variable in the risk estimates. Some studies indicate that the predictive properties of cholesterol might not be as straightforward as widely assumed. Our aim was to document the strength and validity of total cholesterol as a risk factor for mortality in a well-defined, general Norwegian population without known CVD at baseline.

Methods  We assessed the association of total serum cholesterol with total mortality, as well as mortality from CVD and ischaemic heart disease (IHD), using Cox proportional hazard models. The study population comprises 52 087 Norwegians, aged 20–74, who participated in the Nord-Trøndelag Health Study (HUNT 2, 1995–1997) and were followed-up on cause-specific mortality for 10 years (510 297 person-years in total).

Results  Among women, cholesterol had an inverse association with all-cause mortality [hazard ratio (HR): 0.94; 95% confidence interval (CI): 0.89–0.99 per 1.0 mmol L−1 increase] as well as CVD mortality (HR: 0.97; 95% CI: 0.88–1.07). The association with IHD mortality (HR: 1.07; 95% CI: 0.92–1.24) was not linear but seemed to follow a ‘U-shaped’ curve, with the highest mortality <5.0 and ≥7.0 mmol L−1. Among men, the association of cholesterol with mortality from CVD (HR: 1.06; 95% CI: 0.98–1.15) and in total (HR: 0.98; 95% CI: 0.93–1.03) followed a ‘U-shaped’ pattern.

Conclusion  Our study provides an updated epidemiological indication of possible errors in the CVD risk algorithms of many clinical guidelines. If our findings are generalizable, clinical and public health recommendations regarding the ‘dangers’ of cholesterol should be revised. This is especially true for women, for whom moderately elevated cholesterol (by current standards) may prove to be not only harmless but even beneficial.

Ancillary