Risk factor thresholds: their existence under scrutinyBMJ 2002; 324 doi: https://doi.org/10.1136/bmj.324.7353.1570 (Published 29 June 2002) Cite this as: BMJ 2002;324:1570
All rapid responses
Law and Wald make the case for addressing all risk factors,
according to the Absolute Risk of the condition in question.
An otherwise masterly statement from the population
perspective is marred by some overstatements.
1. 'Dose-response relations have no threshold and yield
By logarithmic transformation, most curves can be
straightened out, and many a 'threshold' thereby obscured.
In the Lancet this same week, Shepherd (Lancet 2002; 359: 2271-73) shows how 'the
first cut is the deepest' as regards to cholesterol-lowering, and that attempts
to drive cholesterol further downward exhibits 'diminishing returns'.
The 'proportional reduction' of Law & Wald's own
logarithmically-transformed graphs shows that lowering a diastolic BP of 110 by
10mmHg produces twice as much benefit as lowering a BP of 90mmHg.
To clinicians and patients, as well as resource-managers,
2. 'Treat anyone at high risk'
Blood pressure lowering drugs should not be limited to
people with high blood pressure, nor cholesterol lowering drugs to people with
high serum cholesterol concentrations.
Whilst I agree with the thrust of the argument that we
should target those at high-risk rather than those with a high risk-factor, we
have to recognise that without a sufficient amount of 'reversible risk-factors'
we have nothing to treat !
Hence the need for the later qualification 'Aim for a large
change to achieve a large benefit' which again directs therapy back towards
high levels of reversible risk-factors
Furthermore, current CHD-preventative treatments fall into
two categories : Those modifying a known risk-factor, and those which are not
(eg: Aspirin, or Warfarin). Aspirin and
Warfarin can be aimed at everyone at high-risk, whether or not they have known
3. 'Carry out only necessary randomised trials'
Further Randomised trials ARE necessary to establish
reductions in risk from successively lower starting values.
The claim that 'The evidence from large cohort studies,
however, makes this unnecessary' is a logical mistake. Cohort studies do not
take individuals at high-risk, subject them to potentially toxic treatment, and
record what outcomes occur !
Having constructed a mythical hybrid beast in Fig 4, Law
& Wald despatch their chimaera as implausible.
A more plausible graphical hypothesis is as follows..
^ | x
T | x
C | x
O | x
M | x
E | x
Risk Factor or
Treatment 'Dose' >
It is precisely to identify the 'o' point, where treatment
benefit exceeds treatment risk, that we need additional Randomised trials.
Examples of this are
Treatment benefit is yet discernible below 80-90 diastolic, and between 90-100
diastolic only those at high-risk can be shown to benefit.
b) Atrial Fibrillation:
Anticoagulation in Chronic
AF is indicated only where thrombo-embolism avoided (xxx) exceeds haemorrhage
4. 'Thresholds have been excluded ..
..down to cholesterol levels (3.8 mmol/l)26 and blood
pressure levels (118 systolic/76 diastolic)7 that are close to the prehistoric
levels (table 3).
.. There must be lower limits to the physiological variables
we have considered, beyond which harm will arise: everyone needs blood
pressure, and cholesterol is essential for life. These lower limits are,
however, beyond Western values and not reached by current dietary or drug
But thresholds of risk-benefit have been shown - eg:
Warfarin in AF and anti-hypertensives in mild hypertension..
80 mildly hypertensive (diastolic 90-110) men need to be treated for ten years for
just one of them to avoid a stroke.
Clinicians and Epidemiologists can agree the main point -
'the selection of
individuals for preventive treatment should be based [only] on a person's
absolute level of risk'
if we drop the word [only].
Dr L S Lewis
Competing interests: Law and Wald make the case for addressing all risk factors,according to the Absolute Risk of the condition in question.
on the basis of 45-year-long clinical experience, I agree completely with
Law MR. and Wald NJ (1), who rightly states that “They are not direct
environmental causes of disease, like smoking, but they may be seen as
biochemical or biophysical variables, under partial genetic control, that
are intermediates between environmental factors and disease itself”.
Please, See my site HONCode ID. N°233736,
Constitutions, as well as the Page, I hold weekly in the italian site
www.katamed.it, Semeiotica Biofisica, the article N° 13 “Oncological
Terrain” (2, 3, 4). As a matter of fact, smokers without “oncological
constitution”, i.e. Oncological Terrain, are and will surely not be
involved by any cancer, as allows us to state a large number of clinical
Certainly, interventions to lower blood pressure, serum cholesterol, and
other risk factors [such as IIR (!)] reduce the risk of cardiovascular
disease regardless of initial levels, even reducing them as much as
possible (1). However, in my opinion, we must go “beyond the risk
factors”. In fact, Primary Prevention of the most common and dangerous
human pathologies, e.g. malignancies, depends clearly by easy and quick
bed-side detecting individuals at "real" risk, i.e. involved by well-
defined biophysical-semeiotic constitution, assessed clinically in
“quantitative” way (5). In order to define clinically a “particular”
constitution, which does not exclude the presence of others, of course, it
is necessary to think over the current possibility of gathering at the bed
-side biophysical-semeiotic data, rich of biological and molecular-
biological information on the various human organs, tissues and biological
systems, so that doctor can describe numerous types of biophysical-
semeiotic constitutions, even from the quantitative point of view. Without
any doubt, these data can not be observed at all by the aid of traditional
physic semeiotics, unable of carrying molecular-biological events to
clinical dimension, which really represents the most original and fertile
aspect of Biophysical Semeiotics. Not to speak, obviously, of
sophysticated semeiotics, despensive and unable to be performed on very
large scale. I am sure that going “beyond the risk factors” represents a
new medical "Weltanschauung", which, therefore, needs open-minded
physicians at the "peer-reviews", nowadays, unfortunately, very seldom to
be find, even in "my" dear BMJ.
1) Law MR., Wald NJ. Risk factor thresholds: their existence under
scrutiny. BMJ 2002;324:1570-1576 ( 29 June ).
2) Stagnaro S., West PJ., Hu FB., Manson JE., Willett WC. Diet and Risk of
Type 2 Diabetes. N Engl J Med. 2002 Jan 24;346(4):297-298. [PubMed
–indexed for MEDLINE].
3) Stagnaro-Neri M., Stagnaro S., La “Costituzione Colelitiasica”: ICAEM-
a, Sindrome di Reaven variante e Ipotonia-Ipocinesia delle vie biliari.
Atti. XII Settim. It. Dietol. ed Epatol. 20, 239, 1993.
4) Stagnaro-Neri M., Stagnaro S., Diagnosi Clinica Precoce
dell’Osteoporosi con la Percussione Ascoltata. Clin.Ter. 137, 21-27, 1991
[Pub-Med indexed for MEDLINE] .
5) Stagnaro-Neri M., Stagnaro S., Cancro della mammella: prevenzione
primaria e diagnosi precoce con la percussione ascoltata. Gazz. Med. It. –
Arch. Sc. Med. 152, 447, 1993.
Competing interests: No competing interests