Skepticism about science and medicine

In search of disinterested science

Posts Tagged ‘invalid biomarkers’

Everyone is sick?

Posted by Henry Bauer on 2013/03/13

I might well have titled this post not “Everyone is sick” but “The ‘health-care’ business is sick and sickening”. Healthy people are being told that they are ill, and they are being treated as though they were ill: they are fed physiologically powerful substances that are toxic to various degrees as well as doubtfully beneficial.

There are two interconnected reasons for this:
Natural conditions, particularly those associated with aging (Seeking Immortality? Challenging the drug-based medical paradigm), are pronounced to be sub-optimal (“unhealthy”, “an illness”) and capable of being improved (“treated”, even “cured”).
Illness is judged not by felt or observable symptoms but by biomarkers: things measured by instruments and presumed to reflect some state of ill health, for example blood pressure.

Biomarkers have proliferated as research turns up more and more physiological variables that can be measured. Perhaps the earliest to come into general use was blood pressure. Jeremy Greene has described in fascinating detail the story (beginning in the late 1950s) of measuring blood pressure and thereupon defining and treating hypertension. Shortly thereafter came blood-sugar measuring and the defining and treating of diabetes. Not long after that, measuring cholesterol levels and seeking ways to lower them (Prescribing by Numbers: Drugs and the Definition of Disease, Johns Hopkins University Press, 2007).

The general presumption is that “high” blood pressure, hypertension, is harmful; that it is itself — or at least causes — cardiovascular disease (CVD); that lowering high blood pressure is beneficial to health.
Similarly, it is presumed that elevated levels of blood sugar indicate a tendency to diabetes, and that this can be staved off or even reversed by lowering blood-sugar levels, and that this improves health and potential life-span.
Again, it is presumed that high levels of cholesterol indicate a tendency for plaques to form inside the arteries, and that it is beneficial to health to lower cholesterol levels in the blood.

We are so accustomed to this line of thinking that it may seem remarkable to discover that the evidence in favor of these presumptions is slim to none, and that practices based on those presumptions may be causing harm rather than aiding health. Nevertheless, that’s what the evidence is.

The Institute of Medicine Report, Evaluation of Biomarkers and Surrogate Endpoints in Chronic Disease (IOM 2010), finds that none of the commonly used biomarkers is a valid measure of the illness it supposedly tracks. As to subsequent treatment, Järvinen et al. have pointed out that “There are no valid data on the effectiveness . . . [of] statins, antihypertensives, and bisphosphanates” (the last, e.g. Fosamax, are prescribed against osteoporosis) — British Medical Journal, 342 (2011) doi: 10.1136/bmj.d2175.
That last quote is surely an astonishing assertion, given that innumerable individuals are being fed statins and blood-pressure drugs and bisphosphanates not because they feel ill in any way but purely on the basis of levels of biomarkers (bone density in the case of bisphosphanates).

That Institute of Medicine Report can be downloaded free at the National Academies Press website, or purchased as a volume of some 336 pages (ISBN 978-0-309-15129-0; ed. Christine M. Micheel & John R. Ball; 2010). The provenance seems as authoritative as one can get, since it was prepared for the Committee on Qualification of Biomarkers and Surrogate Endpoints in Chronic Disease; Board on Health Care Services; Board on Health Sciences Policy; Food and Nutrition Board, under the auspices of the Institute of Medicine, a division of the National Academies, “Advisers to the Nation on Science, Engineering, and Medicine”.
Possibly to demonstrate expertise and authority, the Report is not welcoming to readers, especially lay readers; it is virtually unreadable: unfocused, repetitive, altogether badly organized, replete with jargon. The conclusions of greatest interest to potential patients are neither concisely nor clearly stated; they are best divined by thinking about the implications of the various sets of Recommendations. For example:

“Recommendation 5:
5a. Congress should strengthen the FDA’s [Food and Drug Administration] authority to request and enforce postmarket surveillance across drugs, devices, and biologics when approvals are initially based on putative surrogate endpoint data.
5b. Congress should grant the FDA authority to request studies and sufficient authority to act on the results of studies on consumer understanding of claims on foods and supplements.”

What 5a refers to is that drugs and devices are approved on the basis of inadequate trials, so that safety and efficacy could be discovered only after sufficiently large numbers of patients have been exposed to them, but there is no system for monitoring actual performance of drugs and devices once they have been approved. Congress legislated that drug companies must pay the FDA the costs of evaluating requests for initial approval of drugs, but Congress also forbade the FDA from using those funds to monitor performance of the drugs after they have gone into general use. So the possibly fatal “side” effects of drugs become known to the FDA and the public more or less as a matter of chance. There are many examples where knowledge of harm, and later withdrawal off drugs, came only years after some unknown number of individuals had suffered damage, sometimes death. Thus Bayer had their statin (cerivastatin, Baycol) approved in June 1997. Within 4 months, Bayer knew that it could cause rhabdomyolysis — muscle wasting and death. But this risk was not noted in the drug’s package insert for another year-and-a-half, and the drug was withdrawn only in August 2001 (IOM 2010: 205-6), after some hundreds of acknowledged deaths and as many as 100,000 estimated or possible deaths.

What 5b means is that the Food and Drug Administration cannot presently safeguard consumers against unwarranted health claims on food packaging, in some part because the validity of the claims is not known. FDA doesn’t have the resources to do the needed studies itself, and has to rely on the product manufacturers for information.

“Recommendation 6:
6a. The U.S. Department of Health and Human Services should facilitate a coordinated, department-wide effort to encourage the collection and sharing of data about biomarkers for all uses, including drugs, biologics, devices, and foods.
6b. The FDA in coordination with other federal agencies should build needed data infrastructure and surveillance systems to handle the information necessary to gain sufficient understanding of the effects of biomarker use.”

In other words: Even though biomarkers are used universally as measures of health and illness and general physiological status and action, it is not known whether these practices are valid.

In fact, they are known to be NOT valid when employed in the manner most relevant to patients. IOM 2010 points out that biomarkers are used in several ways, for example as an initial screening for potentially useful drugs. Biomarkers are known to be associated in some way with certain conditions, so if a potential medication has no effect on a pertinent biomarker, then it is highly unlikely to be of use in treating that condition. There is nothing to criticize about such a use, nor about research on biomarkers. What does matter profoundly is when a biomarker is used to diagnose illness and treatment is based on that biomarker.

All that’s generally known about biomarkers is that they are associated in some way, correlated, with some physiological condition. For example, people with higher blood pressure have a higher mortality. But as with all statistical associations, this does not demonstrate cause and effect, it does not demonstrate that high blood pressure causes earlier death. It might be that there is some underlying factor that causes both higher mortality and higher blood pressure. In fact there is: increasing age brings both higher blood pressure and higher mortality.

Burrowing into IOM 2010 reveals that none of the commonly used biomarkers are valid for diagnosis of the condition that they are in practice used to diagnose. For example:
The size of cancer tumors is not a measure of how far the disease has advanced or what the prognosis is.
None of the commonly used biomarkers for heart disease are valid indicators of heart disease: not C-reactive protein, not troponin, not overall cholesterol or LDL (“bad” cholesterol) or HDL (“good” cholesterol) — all those are statistically correlated with cardiovascular disease and mortality, but none of them are correlated 100% with those and none of them is indicated to be a cause of heart disease or a contributor to heart disease.

In common parlance, biomarkers are often referred to as “risk factors”. Risk factors are nothing more than statistical correlations, they are not actual risks or causes of disease. Thus for CVD, the traditional risk factors are “smoking, hyperlipidemia, high blood pressure, and diabetes mellitus” (IOM 2101: 142), but those have not been shown to be causally related to CVD. For pertinent comparison: Smoking was shown to be a cause of lung cancer not because of the statistical correlation between smoking and lung cancer but only after animals forced to inhale tobacco smoke were found to develop lung cancer. No such direct test of CVD biomarkers is available.

Drugs administered to treat “hypertension” and “high” cholesterol
have not been proven to be of benefit
in preventing heart disease or to decrease mortality.

Similarly with other biomarkers and treatment based on them.

That is concluded and stated in the most authoritative official literature.


Posted in medical practices, science policy | Tagged: , | 1 Comment »

%d bloggers like this: