Antihypertensive treatment and risk of cardiovascular complications: is the cure worse than the disease?

No! antihypertensive treatment is not worse than the disease! Effective antihypertensive treatment reduces cardiovascular morbility and mortality substantially 1. So why would anybody ask this kind of question? For more than two decades in a number of observational studies 2–6 in the general population, patients on antihypertensive medication were found to have worse prognoses than individuals with hypertension, but without treatment. This was for example already shown by Lars Lindholm and coworkers in 1984 2. Lars Lindholm wisely concluded, that ‘although blood pressure was reduced in accordance with what was internationally accepted, the values were still at least 16/11 mmHg higher than those of the controls. A further reduction of this magnitude, and proper intervention with factors linked to obesity may be one way to further reduce the excess risk in hypertensives’. Randomized controlled trials usually run for a limited number of years, often no more than 3–5 years. It is important to know whether the benefits seen are the same in everyday medical practice for much longer periods. In 1994 a Norwegian paper 3 based on a large population study concluded that the risk for coronary heart disease was higher in treated hypertensives than in untreated, and a very provocative conclusion was allowed by this journal: ‘The beneficial experience from the trials turned into an adverse effect of treatment in the population setting’! Of course this kind of statement was not left undebated 7,8. The most likely explanations for the findings are unsuccessful control of blood pressure 9,10 and a number of other risk factors, as well as so called ‘confounding by indication’ 11,12 that is, those with severe hypertension plus other risk factors are more likely to receive treatment as compared with hypertensives at lower risk. You cannot blame it on the treatment! Now, like a deja vu, in this issue of the journal Asayama and coworkers 13 once more discuss the problem based on a comprehensive Japanese database, with a longitudinal follow up for 9.5 years. The results show that individuals on antihypertensive medications had a higher stroke risk compared with those without treatment for a given blood pressure level, after adjustment for major confounding factors. According to baseline BP readings, subjects were classified into six blood pressure levels, based on the 2007 European Society of Hypertension/European Society of Cardiology guidelines 1, into optimal BP (<120/80 mmHg), normal (120/80–129/84 mmHg), high normal (130/85–139/89 mmHg), grade I hypertension (140/90–159/99 mmHg), grade II hypertension (160/100–179/109 mmHg) and grade III hypertension (≥180/110 mmHg). Next they were classified according to whether they were receiving antihypertensive medications or not. The results show that, for each blood pressure stratum, the risk of a stroke was higher in treated compared with untreated hypertensives, also when the results were adjusted for a number of classical risk factors, including baseline blood pressure. The findings were the same for men and women, although when adjusted for baseline blood pressure the increase in risk of a stroke was no longer significant for women. It is noteworthy that in the untreated individuals there was a stepwise increase in relative hazard for stroke from optimal to grade III hypertension, whereas such a graded risk was not seen in the treated individual. The authors are very careful in the way they discuss the results, and they are well aware of a number of limitations. The results are based on baseline blood pressure and knowledge about antihypertensive treatment at baseline. There is no information about the level of blood pressure and the antihypertensive treatment during the mean follow up period of 9.5 years. Many important changes might have taken place in terms of antihypertensive treatment/compliance with treatment across 9.5 years of follow up. So what are the explanations for the findings in the present study and earlier published ones 2–6,9–11? It has been well established that antihypertensive treatment reduces cardiovascular morbility and mortality in a huge number of randomized controlled trials, with more than 40% reduction in stroke rate for a difference in blood pressure among the treated vs. the placebo group of approximately 10–12/5–6 mmHg 1,14. Furthermore, observational epidemiological data have shown that each time blood pressure is reduced by 20/10 mmHg, the risk for a stroke is halved and this goes down to approximately 115 mmHg in systolic blood pressure 15. Is it the blood pressure? In earlier published population studies there was a marked difference in the blood pressure level between the treated vs. untreated hypertensive individuals, often about 10 mmHg higher or even more in the treated hypertensives 2,5,6,9,10. This indicates that severe hypertensives at high risk receive treatment, but they are unsuccessfully controlled. This is most likely to be the most important explanation. In the present study, even though the patients were grouped in six different strata of blood pressure levels, there is still room for blood pressure differences. The mean values of systolic/diastolic blood pressure within the six groups for untreated vs. treated are quite similar, but on the average for all the individuals' systolic/diastolic blood pressure is 7/6 mm higher in the treated individuals as compared with untreated, and as mentioned in women the increased risk was no longer significant when adjusted for blood pressure levels. So very likely in the former and the present study higher blood pressure in treated individuals explains a significant part of the increased risk. But certainly, there is an excess residual risk that is not easily recognized, managed, or both. Is it the choice of antihypertensive treatment? There is no doubt that the major benefit linked to blood pressure lowering is normalization of blood pressure 1,14. On the contrary, recent meta-analyses indicate that the benefit in terms of risk reduction is less when β blockade is used as a routine first line treatment 16,17); this is especially true for atenolol and maybe to a lesser extent for metoprolol 16. This is when compared with placebo, but also with other forms of active treatment, such as calcium blockers +/− inhibitors of the renin–angiotensin–aldosterone system. Whether new β blockers with vasodilating properties, such as carvedilol and nebivilol, possess benefits above traditional β blockers is not known. Beta-blockers, to some extent thiazides and definitely the combination, provoke new onset type 2 diabetes 14,18 – or disclose diabetes earlier than otherwise. This negative effect might have an unfavourable impact on cardiovascular outcome 19: although, as of today, not so well proven 18. In the Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHATT) study α-blockade by means of doxazocin led to increased heart failure rate 20, but it has been discussed whether the endpoint heart failure was well adjudicated 14. More than a decade ago it was intensively postulated, based on observational studies, that Ca-blockers negatively influenced cardiovascular mortality 21. Solid evidence in favour of Ca-blockers, such as amlodipine, from controlled trials 22,23, has effectively closed the issue. It is a standing debate as to whether ACE-inhibitors/angiotensin receptor blockers confer benefits beyond blood pressure reduction in special subgroups. The discussion is not quite settled yet 14,24. Does a J-shaped curve come into play? It has been quite a long discussion whether there is a J-shaped relationship between achieved blood pressure and risk for cardiovascular diseases. The discussion has especially been related to patients with coronary heart disease and whether too pronounced a fall in blood pressure leads to increased mortality 25. However, a J-point is well below a systolic blood pressure of 115 mmHg 15, and well below the blood pressure levels usually seen in this type of observational study 26, including the present one. Is the residual, excess risk in treated hypertensives explained by other cardiovascular risk factors than high blood pressure? Most likely, antihypertensive treatment per se is a marker, not only for the severity of hypertension, but also for other cardiovascular risk factors such as smoking, hypercholesterolaemia and diabetes mellitus that lead to a high risk for cardiovascular events 11. Even though hard to prove, doctors that initiate antihypertensive treatment comprehend/recognize an individual of increased global risk, not just related to the level of blood pressure 27. There might be several residual confounding factors in addition to classical risk factors, which means that stroke incidence, even adjusted for classical risk factors, including age, would be high among treated individuals. The issue was addressed by Klungel et al.11 who carefully matched the treated patients with the untreated control group for a number of risk factors. By this approach they did not find any difference in stroke risk among treated versus untreated individuals. This explanation is also offered by Asayama and coworkers 13, so called ‘confounding by indication’. The issue points to the importance of meticulous risk factor assessment in hypertensive individuals, taking classical risk factors and presence of target organ damage into account. This can be done by means of a risk chart or the risk assessment proposed by the ESH/ESC guidelines 1. A single, very robust integrated measure of risk is tiny amounts of albuminuria, not just at baseline, but also during intervention because changes in albuminuria during intervention translate to changes in risk of cardiovascular diseases 28. Conclusion When treated hypertensive patients still are at increased risk for cardiovascular morbidity and mortality the most important explanation is lack of control of blood pressure and other risk factors. As long as only half of treated patients reach the treatment goals – and by no means normal/optimal blood pressure – there is a huge potential for further prevention of cardiovascular complications. There is a need for intensified combination treatment, including a greater use of fixed-dose-combination agents. Careful assessment and intervention directed towards global risk, including target organ damage, is vitally important. As stated by Asayama and coworkers 13: ‘Healthcare providers need to be vigilant for residual cardiovascular risk in treated hypertensive patients’.