Skip to main content
NCBI home page
The BMJ logoLink to The BMJ
. 2002 Nov 9;325(7372):1114.

Risk factor thresholds

Threshold is £37 000 per QALY

Michael A Soljak 1
PMCID: PMC1124591  PMID: 12424179

Editor—Although I agree with most of Law's and Wald's conclusions with regard to risk factors, I cannot agree that, as a result, treatment thresholds do not exist.1 With regard to the risk of coronary heart disease, the recent joint British recommendations recommend starting treatment of high blood pressure at an absolute 10 year risk of coronary heart disease of 15%, and of a high lipid ratio at 30%.2 Neither these recommendations nor those of the Standing Medical Advisory Committee explain why these particular thresholds have been set. (Neither do they mention when treatment should be stopped.) But could or should it have something to do with cost? I think that, despite the article's title, Law and Wald acknowledge this implicitly by saying that people at high risk should be targeted.

There have been several published cost effectiveness analyses of lipid lowering drugs. The report from Pickin et al puts the cost per (presumably good quality) year of life gained of treating coronary heart disease risk above 3% per year at £8200, which they describe as of comparable cost effectiveness to many treatments in wide use.3 They say, however, that treatment below this level is unlikely to be affordable. The de facto threshold currently being used by the National Institute of Clinical Excellence is considerably higher—about £37 000 per QALY.

Ethical questions such as the value the NHS and other health systems should place on preventive rather than immediately lifesaving care remain largely undiscussed. Perhaps that is why so many authors overlook that resources are scarce and so thresholds must always exist. Isn't it time that this collective blind spot was removed?

References

  • 1.Law MR, Wald NJ. Risk factor thresholds: their existence under scrutiny. BMJ. 2002;324:1570–1576. doi: 10.1136/bmj.324.7353.1570. . (29 June.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.British Cardiac Society; British Hyperlipidaemia Association; British Hypertension Society; British Diabetic Association. Joint British recommendations on prevention of CHD in clinical practice. Heart. 1998;80(suppl 2):S1–29. [PMC free article] [PubMed] [Google Scholar]
  • 3.Pickin DM, McCabe CJ, Ramsay LE, Payne N, Haq IU, Yeo WW, et al. Cost effectiveness of HMG-CoA reductase inhibitor (statin) treatment related to the risk of CHD and cost of drug treatment. Heart. 1999;82:325–332. doi: 10.1136/hrt.82.3.325. [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 2002 Nov 9;325(7372):1114.

Hypothesis is dangerous nonsense

Malcolm E Kendrick 1

Editor—Law and Wald discuss risk factor thresholds,1-1 having established in a previous paper on the time lag hypothesis that the risk factors in French men aged 45-641-2 are:

  • Total cholesterol concentration 6.1 mmol/l

  • Systolic blood pressure (their previous “best blood pressure risk factor”) 150 mm Hg

  • High density lipoprotein concentration 1.3 mmol/l

  • Smoking 32%

Add to this data from the British Heart Foundation:

  • Average body mass index 26.6

  • Those never exercising 32%

  • Rate of coronary heart disease 128/100 000/year

The risk factors in British men aged 45-64 are:

  • Total cholesterol concentration 6.2 mmol/l

  • Systolic blood pressure 148 mm Hg

  • High density lipoprotein concentration 1.3 mmol/l

  • Smoking 29%

  • Average body mass index 26.6

  • Those never exercising 24%

  • Rate of coronary heart disease 487/100 000/year

Can Law and Wald fit these figures onto their semilogarithmic scale? The suggestion that no levels of any risk factor in the Western world are currently normal, and that what we call a normal blood pressure is actually high and should be lowered, is dangerous nonsense.

Figure.

Figure

Open in a new tab

J shaped curve of total mortality v total cholesterol concentration based on data from Jacobs et al, Circulation 1992;86:1046-60.

Are Law and Wald aware of data from Framingham, which show that falling cholesterol concentrations are directly associated with an increased risk of coronary heart disease?1-3 Are they aware of research from Japan that shows a completely inverse relation between rising cholesterol concentrations and deaths from coronary heart disease?1-4 Hundreds of papers contradict the association between raised cholesterol concentrations and death from coronary heart disease.

Shestov in his lipid clinics study in Russia even showed an inverse relation, with higher rates of coronary heart disease in patients with hypocholesterolaemia. The Honolulu study shows that, in people older than 50, a low cholesterol concentration is by far the most important risk factor for premature death.1-5 Law and Wald did not show one curve relating to cholesterol lowering—the J shaped curve of total mortality with 5.2 mmol/l at the bottom of that curve (figure).

Law and Wald are effectively suggesting that there is no non-dangerous blood pressure or cholesterol concentration and that, therefore, almost everyone in the Western world should be given some kind of drug treatment. This is dangerous nonsense, and we should not be afraid to say so.

References

  • 1-1.Law MR, Wald NJ. Risk factor thresholds: their existence under scrutiny. BMJ. 2002;324:1570–1576. doi: 10.1136/bmj.324.7353.1570. . (29 June.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 1-2.Law MR, Wald NJ. Why heart disease mortality is low in France: the time lag explanation [with commentaries by M Stampfer, E Rimm, D J P Barker, J P Mackenbach, and A E Kunst] BMJ. 1999;318:1471–1480. doi: 10.1136/bmj.318.7196.1471. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 1-3.Anderson KM, Castelli WP, Levy D. Cholesterol and mortality: 30 years of follow-up from the Framingham study. JAMA. 1987;257:176–180. doi: 10.1001/jama.257.16.2176. [DOI] [PubMed] [Google Scholar]
  • 1-4.Okayama A, Ueshima H, Marmot MG, Nakamura M, Kita Y, Yamakawa M. Changes in total serum cholesterol and other risk factors for cardiovascular disease in Japan 1980-1989. Int J Epidemiol. 1993;22:1038–1047. doi: 10.1093/ije/22.6.1038. [DOI] [PubMed] [Google Scholar]
  • 1-5.Schatz IJ, Masaki K, Yano K, Chen R, Rodriguez BL, Curb JD. Cholesterol and all-cause mortality in elderly people from the Honolulu Heart Program: a cohort study. Lancet. 2001;358:351–355. doi: 10.1016/S0140-6736(01)05553-2. [DOI] [PubMed] [Google Scholar]
BMJ. 2002 Nov 9;325(7372):1114.

Pharmacological treatment should not be determined by age alone

Steven M Laitner 1

Editor—Law and Wald provide evidence for the absence of thresholds in the relation between risk factor and disease.2-1 They conclude that people with high absolute risk will benefit from a reduction in risk factors whatever their initial risk factor level. They say that in people without cardiovascular disease, intervention to change risk factors could be introduced when a person's risk of a disease event over the following few years exceeds a specified value. Risk could be estimated from age alone or age and sex. Individuals at high risk should receive drug treatment to modify all important reversible risk factors simultaneously.

Although combination pharmacological cardiovascular risk factor reduction for the whole elderly population represents an ideal strategy for the pharmaceutical industry it has numerous limitations as a public health strategy.

(1) By setting an absolute risk threshold for treatment (determined by age only) the proposed strategy still targets only those at high risk of disease (the tail of the population distribution curve of risk factors) and therefore can only have a minimal effect on the overall burden of disease in the population.

(2) There will be major opportunity costs in treating the whole elderly population with combination drugs and consequently other preventive, treatment, care and rehabilitation services for elderly people would be constrained.

(3) The authors' statement that lower limits of thresholds (such as blood pressure), beyond which harm will arise, are not reached by current drug treatment is false—the risks of polypharmacy in elderly people are significant2-2,2-3 and would be increased.

(4) Even in elderly people, absolute risk of coronary heart disease may not reach 3% per year without additional risk factors such as smoking and diabetes. According to the Framingham coronary risk prediction score (www.nhlbi.nih.gov/about/framingham/risktmen.pdf), a man aged 70-74, total cholesterol 5.18-6.21 mmol/l, high density lipoprotein 1.17-1.29 mmol/l, and blood pressure (140-159)/(90-99) has a risk of coronary heart disease of 2.5% per year; for a woman the risk is only 1.3%.

Targeting people who have had a vascular event or who have diabetes with treatment to reduce risk factors is an appropriate “high risk” strategy. The only appropriate strategy, as challenging as it may be, to reduce the risk of vascular disease in the rest of the population (who cause most of the burden of disease) is to reduce the average levels of risk factors in the population, through the promotion of a healthy diet, exercise, and smoking cessation. Proposals to use drug treatment for primary prevention in the whole population over a certain age should be resisted.

References

  • 2-1.Law MR, Wald NJ. Risk factor thresholds: their existence under scrutiny. BMJ. 2002;324:1570–1576. doi: 10.1136/bmj.324.7353.1570. . (29 June.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2-2.Beyth RJ, Shorr RI. Epidemiology of adverse drug reactions in the elderly by drug class. Drugs Aging. 1999;14:231–239. doi: 10.2165/00002512-199914030-00005. [DOI] [PubMed] [Google Scholar]
  • 2-3.Malhotra S, Karan RS, Pandhi P, Jain S. Drug related medical emergencies in the elderly: role of adverse drug reactions and non-compliance. Postgrad Med J. 2001;77:703–707. doi: 10.1136/pmj.77.913.703. [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 2002 Nov 9;325(7372):1114.

Authors' reply

Malcolm Law 1,2, Nicholas J Wald 1,2

Editor—We pointed out that within the range of values of common risk factors for disease in our population there is no evidence of a threshold below which further modification of risk factors yields no further reduction in the risk of disease. None of the correspondents has provided any evidence to the contrary.

Soljak accepts the absence of biological thresholds but mentions other thresholds such as the costs per year of life gained. With low cost drugs (many useful drugs have come off patent or are shortly to come off patent) such financial thresholds could largely disappear.

We believe that Kendrick's views are incorrect and have been refuted in detail elsewhere.3-1,3-2 The J shaped curve relating to total mortality shows a combination of two different processes. The risk of occlusive cardiovascular disease increases with increasing serum cholesterol concentration, but many fatal chronic diseases lower serum cholesterol concentration in their early stages.3-2

Laitner believes that there are limitations to our view that reduction of risk factors needs to be offered on a much wider basis using drugs if necessary. Of course, the preventive treatment would have to be effective, reasonably safe, inexpensive, and sufficiently simple for people to take widely without frequent medical examination or monitoring. We agree that the assessment of such a preparation would have to be undertaken in elderly people and a quantitative assessment of benefits and risks determined. As Laitner points out, selecting a cut-off point for absolute risk such as 3% per year (a very high risk) means that many who will die of cardiovascular disease would be below the risk threshold. This is the main reason for simply using age as the selection criterion.

The important point in our paper is that there are no biological thresholds that should limit the extent to which risk factors are reduced in the community, whether through drug treatment or diet.

References

  • 3-1.Heart Protection Study Collaborative Group. MRC/BHF Heart protection study of cholesterol lowering with simvastatin in 20 536 high-risk individuals: a randomised placebo-controlled trial. Lancet. 2002;360:7–22. [Google Scholar]
  • 3-2.Law MR, Thompson SG, Wald NJ. Serum cholesterol reduction and health: assessing possible hazards. BMJ. 1994;308:373–379. doi: 10.1136/bmj.308.6925.373. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from BMJ : British Medical Journal are provided here courtesy of BMJ Publishing Group

RESOURCES