THIRTY YEARS AGO and already well into his career, the Oxford epidemiologist Sir Richard Peto wondered - with characteristic restraint – if any of his work had ever contributed to the avoidance of a single unnecessary death. Today, honoured for his studies throughout the world, he should have no such doubts. Mortality trends in many developed countries now show steep and consistent declines in premature death from
heart disease and stroke, and in the UK vascular mortality rates in middle age have, remarkably, fallen by more than half since 1990.
The Clinical Trial Service Unit and Epidemiological Studies Unit of Oxford University (CTSU), of which Peto and Professor Rory Collins have been co-directors for more than 20 years, has made substantial contribution to these trends in its major studies of smoking, blood lipids, blood pressure, aspirin and streptokinase.
Tomorrow, Peto will deliver the Geoffrey Rose Lecture on Population Sciences on “Changing vascular mortality: prevention, treatment and national trends."
The most influential trials so far performed by the CTSU - from a 30-year catalogue of major trials - are the 1988 ISIS-2 trial of antiplatelet and thrombolytic treatment in acute MI, and the Heart Protection Study, a five-year secondary prevention trial in more than 20,000 patients, which showed that daily simvastatin reduced the incidence of MI, stroke and revascularisation by about one-third.
Sir Richard, who even in everyday conversation measures his words carefully, describes the impact of both studies as "enormous", but, as the statistician to these trials, he stresses that the definitive answers came from the meta-analyses of these and all related studies which CTSU helped organise - the Anti-Thrombotic Trialists’ (ATT), Fibrinolytic Treatment Trialists’ (FTT), and the Cholesterol Treatment Trialists’ (CTT) collaborations. "When we put all the trial results together," says Peto, "CTT showed absolutely conclusively that the benefits of statins are substantial and the hazards small."
Today, high-impact non-communicable disease epidemiology is securely embedded in Oxford. When the late Sir Richard Doll (who had helped demonstrate the causative link between smoking and lung cancer) moved from London to Oxford in 1969, he took with him a team of young epidemiologists, including Peto, who laid the intellectual foundations of the CTSU. Now, in the Unit's utilitarian modern building, the words of Sir Richard Doll still define the key principle that underlies its work: "Death in old age is inevitable, but death before old age is not."
Today, Peto and colleagues emphasise the need for large-scale observational and randomised evidence in the prevention and treatment of major diseases. In their metaanalyses they mistrust “random effects” models, preferring instead the "weighted average" of all study results. Indeed, it was Peto's group in the 1970s which introduced the concept of metaanalysis in vascular disease, and with it the realisation that even a modest effect in a widely prevalent condition could have a dramatic public health impact.
But the downside of big trials today is their growing burden of administration. "I think research is over-regulated," says Peto. "The European Clinical Trials Directive made studies much more complicated, and much more expensive. That’s very damaging. The fact is that I now spend more than half my time on inappropriate administrative requirements."
But despite the regulation, clear results continue to emerge. "Even in the 1990s," says Peto, "there was still discussion of whether lowering blood pressure or blood cholesterol levels would have any net effect on mortality. Those doubts have now been demolished, simply by the accumulation of overwhelmingly clear evidence."
...and a clear picture of risk estimation
Risk-profiling - as evident in a new primary care prediction tool from the Framingham group or the QRISK score from Britain - is now high on the CVD agenda, and a Symposium this afternoon will ask if we've got risk estimation right yet.
The well established risk factors for heart disease are now embedded in all public prevention programmes. Indeed, a report published online by Circulation in July found that if smoking cessation and five "fairly cost-effective measures" were applied (aspirin for high-risk individuals, control of pre-diabetes, weight loss in the obese, lowering blood pressure in diabetes, and lowering LDL cholesterol in CHD) huge falls in MI and stroke could be achieved.
However, while age, gender, smoking, systolic blood pressure, body mass index, family history, type 2 diabetes, renal disease and rheumatoid arthritis seem now accepted beyond doubt as risk factors, difficulties - even controversy - remain with others; the effects of ethnicity lipoprotein cholesterol and social class – and especially their inter-individual effects - have all been under renewed scrutiny in the past year or so.
Indeed, whatever other controversies the ENHANCE trial threw up, its findings (that reduced levels of LDL cholesterol were not associated with reduced atherosclerosis) were
also at odds with conventional wisdom. Similarly, in June this year, following the application of "metabolic syndrome" criteria to subjects in two prospective trials, only a weak association was found between the syndrome and incident cardiovascular disease. And new data from two previously published studies (IDEAL and EPIC) even suggested that high levels of plasma HDL cholesterol and large HDL particles were associated with an increased risk of coronary artery disease.
There is, however, little doubt that the risk of cardiovascular disease is directly related to social class. For more than four decades studies have consistently shown that the greater an individual's social deprivation, the higher the risk of CVD. Social disadvantage seems to increase CVD mortality in middle-aged men at least twofold, and in women three-fold.