Home or ambulatory blood pressure measurements

Some people develop excessive and unrepresentative blood pressure rises when attending the doctor’s surgery, so-called ‘white coat’ hypertension. These patients can be diagnosed if they use a blood pressure machine themselves at home or by 24-h ambulatory blood pressure monitoring. Home blood pressure measurement is inexpensive but it is important to have a machine of validated accuracy that the patient can use properly. Ambulatory blood pressure monitoring over 24 h is also useful for patients who have unusual variability in blood pressure, resistant hypertension or symptoms suggesting hypotension. Home or ambulatory blood pressure measurements are usually lower than clinic recordings, on average by 12/7 mmHg.

Secondary causes

It is important to take a careful history checking for features that might suggest a possible secondary cause of hypertension. Examples would be symptoms of renal disease, for example, haematuria, polyuria, etc., or the paroxysmal symptoms that suggest the rare diagnosis of phaeochromocytoma and include headache, postural dizziness, syncope. A careful physical examination should be performed for abdominal bruits, which suggest possible renal artery stenosis, radiofemoral delay which suggest coarctation of the aorta and palpable kidneys which suggest polycysctic kidney disease. Laboratory analysis should include a full blood count, electrolytes, urea, creatinine and urinalysis. In some patients, further investigations may be appropriate, for example, ultrasound of the abdomen or isotope renogram where renal disease is suspected. A renin–angiotensin ratio is a useful screening test to investigate for possible hyperaldosteronism while serum metanephrine and urinary catecholamines may detect underlying phaeochromocytoma.

A low serum potassium may alert to the presence of hyperaldosteronism but it should be remembered that renin levels are suppressed by β-blockers and aldosterone by angiotensin converting enzyme inhibitors and receptor antagonists. A very high aldosterone/renin ratio may suggest Conn’s syndrome or primary hyperaldosteronism. This is usually caused by a benign adenoma or simple hyperplasia within the zona glomerulosa of the adrenal gland, the presence of which may be demonstrated by CT or MRI scanning. The tumours may be surgically resected, but where there is a suggestion of hyperaldosteronism and no obvious tumour on imaging, patients may still respond to spironolactone, an aldosterone antagonist, while remaining relatively resistant to other antihypertensives.

Contributing factors

The patient should also be assessed for possible contributory factors to hypertension such as obesity, excess alcohol or salt intake and lack of exercise. Occasionally, hypertension may be provoked by the use of drugs (see Box 19.2), including over-the-counter medicines used as cold and flu remedies. Other risk factors should also be documented and addressed, for example, smoking, diabetes and hyperlipidaemia. It is important to establish whether there is a family history of cardiovascular disease.

Evidence of end-organ damage

The patient should also be examined carefully for evidence of end-organ damage from hypertension. This should include examination of the optic fundi to detect retinal changes. An ECG should be performed to detect left ventricular hypertrophy or subclinical ischaemic heart disease. It is advisable to check the renal function and test the urine for signs of micro-albuminutria which may be an indicator of a higher risk of future end-stage renal disease and overall vascular risk.

Determination of cardiovascular risk

An accurate assessment of cardiovascular disease risk is essential before recommending appropriate management in hypertension. Patients with documented atheromatous vascular disease, for example, previous myocardial infarction or stroke, angina or peripheral vascular disease are at high risk of recurrent events. Those with type 2 diabetes over 40 years of age are also at high risk and can be regarded as ‘coronary equivalents’, that is, with risks similar to non-diabetic patients with previous myocardial infarction. For non-diabetic patients without vascular disease it is necessary to estimate cardiovascular risk (see Chapter 24). A 10-year cardiovascular disease risk of 20% (equivalent to a 15% coronary heart disease risk) is regarded as an appropriate threshold for antihypertensive therapy in patients with moderate hypertension, as well as for lipid-lowering therapy. Treatment decisions based on these tables will favour treatment in elderly subjects. While a younger patient may be at lower absolute risk over 10 years and may not meet the criteria for blood pressure and lipid treatment, they may be at higher lifetime and longer term risk of premature death and vascular disease and, thus, still merit risk factor intervention.

Other factors to consider include microalbuminuria which increases cardiovascular risk by a factor of 2–3 and the combination of reduced GFR and microalbuminuria may increase risk by as much as six-fold (Cirillo et al 2008; Sehestedt et al., 2009).

Treatment thresholds

Treatment thresholds are summarised in Table 19.1. Lifestyle advice should be provided to all patients with any degree of hypertension. Patients with severe hypertension (>220/120 mmHg confirmed on several readings on the same occasion) should be treated immediately and some guidance suggests that dual therapy should be commenced immediately in patients with blood pressure >20 mmHg above their target as monotherapy is unlikely to be fully effective (Mancia et al., 2007). Patients with blood pressures in the range 160–220/100–120 mmHg should be monitored over several weeks and treated if blood pressure remains in this range. The period of observation before starting treatment depends on the severity of the hypertension and the presence or absence of end-organ damage (see Table 19.1). Patients whose blood pressure is in the range 140–159/90–99 mmHg should be observed annually unless they have evidence of target organ damage, cardiovascular complications, diabetes or a calculated cardiovascular risk >20% over 10 years, in which case drug treatment should be offered. Patients with blood pressure in the range 135–139/85–89 mmHg should be reassessed annually, while those with blood pressure lower than this can be rechecked every 5 years.

Table 19.1 Threshold blood pressures for intervention

Target blood pressures

Within the Hypertension Optimal Treatment (HOT) study (Hansson et al., 1998), patients were allocated diastolic target blood pressures of <90, 85, 80 mmHg. The study struggled to stratify patients effectively into these treatment groups but analysis suggested that the optimum target blood pressure was <140/85 mmHg with little benefit in lowering to lower levels of 120/70 mmHg but also little evidence of harm.

The UK Prospective Diabetes Study Group (1998a,b) suggested ‘tight’ blood pressure control was better than less tight in patients with non-insulin-dependent diabetes. The targets in the UK Prospective Diabetes Study were ‘tight’ <150/85 mmHg and ‘less tight’ <180/105 mmHg but the actually achieved blood pressures were lower, for example, 154/87 mmHg versus 144/82 mmHg. Recommendations in diabetics have, therefore, suggested treating to a target of 140/80 mmHg or less, although few studies have successfully lowered blood pressure to these levels.

A more recent study (Cardio-Sis) randomised non-diabetic subjects with systolic blood pressure >150 mmHg to target systolic blood pressure of <140 or <130 mmHg (Verdecchia et al., 2009). The primary end-point was left ventricular hypertrophy though the secondary end-point of a composite cardiovascular end-point was reduced (as well as the primary end-point) in the 130-mmHg group, with no increase in adverse events. This, however, is not robust enough evidence to recommend a reduction in blood pressure target levels and would require a larger study of hard clinical end-points to confirm these findings.

Achievement of target blood pressures is incorporated as a quality indicator for the General Medical Services Contract for primary care doctors in the UK. Diabetic patients are an exception and benefit from more aggressive blood pressure reduction. Target blood pressures for diabetic and non-diabetic subjects are summarised in Table 19.2. It should be emphasised that the audit standard will not be achieved in all patients.

Table 19.2 Target clinic blood pressures according to British Hypertension Society guidelines 2004 (Williams et al., 2004)

β-Adrenoreceptor antagonists

The mode of action of β-adrenoreceptor antagonists in hypertension is uncertain. β-Adrenoreceptor blockade reduces cardiac output in the short term and during exercise. They also reduce renin secretion by antagonising β-receptors in the juxtaglomerular apparatus. Central actions may also be important for some agents. Non-selective β-blockers may give rise to adverse effects as a result of antagonism of β₂-adrenoceptors, that is, asthma and worsened intermittent claudication. However, the so-called ‘cardioselective’ (β₁-selective) β-blockers are not entirely free of these adverse effects. Patients who develop very marked bradycardia and tiredness may tolerate a drug with partial agonist activity such as pindolol.

β-Adrenoreceptor antagonists also have substantial clinical trial evidence of benefit over placebo in hypertension, and are relatively inexpensive. However, their use is declining and they have been relegated to fourth-line therapy in the UK according to NICE guidance (Fig. 19.1). This recommendation largely stems from the evidence that they may be less effective at preventing stroke in conjunction with their diabetogenic effects. The Losartan For Endpoint reduction in hypertension (LIFE) study compared an atenolol/thiazide-based regime with a losartan-based regime and demonstrated equivalent levels of blood pressure reduction but with a small excess incidence of stroke in the atenolol arm (Dahlöf et al., 2002). In the Anglo Scandinavian Cardiac Outcomes Trial (ASCOT) study, the risk of diabetes was 2.5% higher in the atenolol arm compared with the amlodipine arm with similar increased risk of diabetes found within the atenolol arm of the LIFE study (Dahlöf et al., 2005). A Cochrane review warned of the excess risk in developing diabetes in patients prescribed combinations of thiazide diuretics and β-blockers. This would equate to one new case per 500 treated (Mason et al., 2004). The combination of thiazide and a β-blocker should, therefore, be avoided if possible, particularly in those who are at risk of developing diabetes (e.g. obese, strong family history of diabetes, South Asian origin).

Fig. 19.1 Algorithm for drug sequencing in hypertension.

To complicate matters, however, a long-term 20-year follow-up study of the UKPDS study found similar cardiovascular outcomes between patients on β-blockers and ACE inhibitors with a reduction in all causes of mortality which actually favoured β-blockers (Holman et al., 2008). β-Blockers do remain most suitable for younger hypertensives who have another indication for β-blockade, such as coronary heart disease. β-Blockers are also effective in suppressing atrial fibrillation and this may be one group of patients where first-line therapy with β-blockers is still merited.

It can be safely assumed that the place for β-blockers for patients with hypertension is likely to remain controversial.

Diuretics

There is substantial clinical trial evidence that benefit is obtained from the use of thiazide, for example, bendroflumethiazide, hydrochlorothiazide, or thiazide-like, for example, chlortalidone, indapamide, diuretics in hypertension; these drugs are both inexpensive and well tolerated by most patients. Their diuretic action is achieved by blockade of distal renal tubular sodium reabsorption. Initially, they reduce blood pressure by reducing circulating blood volume but in the longer term they reduce total peripheral resistance, suggesting a direct vasodilatory action.

Although generally well tolerated, thiazide and thiazide-like diuretics may cause hypokalaemia, small increases in LDL-cholesterol and triglyceride, and gout associated with impaired urate excretion. Erectile dysfunction is also common.

Most blood pressure lowering occurs with very low doses of thiazide diuretics. Increasing the dose substantially increases the risk of metabolic disturbance without causing further blood pressure reduction. For bendroflumethiazide it is rarely (if ever) appropriate to use doses greater than 2.5 mg/day and a dose of 1.25 mg daily is often effective. Most studies of diuretics have also incorporated β-blockers and this combination can have adverse metabolic consequences which may lead to new onset diabetes. Within the Anti-hypertensive and Lipid Lowering treatment to prevent Heart Attack Trial (ALLHAT), the absolute risk of developing diabetes was 3.5% higher in the chlortalidone group than the lisinopril group (ALLHAT Collaborative Research Group, 2002).

As with the β-blocker saga it remains an issue of contention as to whether this diabetic tendency is clinically significant. There was no reduction in efficacy associated with thiazide use in ALLHAT and indeed there were less heart failure outcomes in the diuretic-treated patients compared with those receiving calcium channel blockers or ACE inhibitors.

Loop diuretics are no more effective at lowering blood pressure than thiazides unless renal function is significantly impaired or the patient is receiving agents that inhibit the renin–angiotensin system. They are also a suitable choice if heart failure is present.

Spironolactone, an aldosterone antagonist, is not suitable for first-line therapy but is an increasingly important treatment option for patients with resistant hypertension. Where hyperaldosteronism is suspected, spironolactone may prove to be effective. Spironolactone is a potassium sparing diuretic and should be used with caution especially if used in combination with ACE inhibitors or angiotensin receptor blockers (ARBs), and should almost always be avoided with other potassium sparing diuretics, for example, amiloride.

Renin-angiotensin-aldosterone antagonists

ACE inhibitors block the conversion of angiotensin I to angiotensin II, while ARBs block the action of angiotensin II at the angiotensin II type 2₁ receptor. Since angiotensin II is a vasoconstrictor and stimulates the release of aldosterone, antagonism results in vasodilation and potassium retention as well as inhibition of salt and water retention. ACE inhibitors also block kininase production and, thus, prevent the breakdown of bradykinin. This appears to be important in the aetiology of ACE inhibitor induced cough, which is a troublesome side effect in 10–20% of users. ARBs do not inhibit kininase and are an appropriate choice for patients who are intolerant of ACE inhibitors because of cough. ACE inhibitors are also associated with a significant incidence of angioedema, which can in severe cases cause dangerous swelling of the pharyngolargyngeal area leading to stridor, threatening the patient’s airway. This adverse reaction is commoner in black subjects.

Calcium channel blockers

These agents block slow calcium channels in the peripheral blood vessels and/or the heart. The dihydropyridine group work almost exclusively on l-type calcium channels in the peripheral arterioles and reduce blood pressure by reducing total peripheral resistance. In contrast, the effect of verapamil and diltiazem are primarily on the heart, reducing heart rate and cardiac output. Long-acting dihydropyridines are preferred because they are more convenient for patients and avoid the large fluctuations in plasma drug concentrations that may be associated with adverse effects.

Although effective for lowering blood pressure and preventing cardiovascular events, adverse effects are common, for example, oedema and flushing. Gum hypertrophy may occur with dihydropyridines and constipation with verapamil. Concerns have previously been raised by observational studies (Psaty et al., 1995) and meta-analysis (Furberg et al., 1995) that there may be an increased risk of coronary heart disease in recipients of dihydropyridine calcium channel blockers. However, randomised clinical trials have not confirmed these observations (Gong et al., 1996, Staessen et al., 1997) and have indicated that dihydropyridines are of similar efficacy to thiazide diuretics in preventing cardiovascular events (Brown et al., 2000).

α-Adrenoreceptor blockers

Drugs of this class antagonise α-adrenoceptors in the blood vessel wall and, thus, prevent noradrenaline (norepinephrine)-induced vasoconstriction. As a result, they reduce total peripheral resistance and blood pressure. Prazosin was originally used but had the disadvantage of being short-acting and causing first-dose hypotension. Newer agents such as doxazosin and terazosin have a longer duration of action. There are concerns about the first-line use of β-blockers since the ALLHAT study has indicated that doxazosin is more often associated with heart failure and stroke than thiazide diuretics (ALLHAT Collaborative Research Group, 2000). However, they are an appropriate choice as add-in therapy for patients inadequately controlled using other agents. They can frequently cause postural hypotension but may alleviate symptoms in men with prostatic hyperterophy.

Centrally acting agents

Methyldopa and moxonidine inhibit sympathetic outflow from the brain, resulting in a reduction in total peripheral resistance. Methyldopa is not widely used because it has pronounced central adverse effects, including tiredness and depression. It continues to be used in pregnancy, since it does not cause fetal abnormalities. It is also occasionally used in patients with resistant hypertension. Moxonidine is a newer agent that blocks central imidazoline and α₂-adrenoceptors found within the medulla oblongata of the brain. It can cause side effects of dry mouth, headache, fatigue and dizziness, although it appears to have fewer central adverse effects than methyldopa. Other centrally acting agents such as clonidine and reserpine are almost never used in modern practice because of their pronounced adverse effects.

Other agents

Several other drugs are available for use for people with more resistant hypertension. Minoxidil is a powerful antihypertensive drug but its use is associated with severe peripheral oedema and reflex tachycardia. It should be restricted to patients with severe hypertension who are also taking β-blockers and diuretics. It causes pronounced hirsutism and is not a suitable treatment for women. Hydralazine can be used as add-on therapy for patients with resistant hypertension but is not well tolerated as it is a profound vasodilator and may occasionally be associated with drug-induced systemic lupus erythematosus. Sodium nitroprusside is a direct-acting arterial and venous dilator that is administered as an intravenous infusion for treating hypertensive emergencies and for the acute control of blood pressure during anaesthesia. Hypertension has previously been treated with ganglion blockers such as guanethidine but these drugs are now of historical interest only.

Recent additions to the licensed armory of antihypertensive agents include the renin antagonist aliskiren. There is evidence that this agent may have a similar blood pressure lowering effect to other agents and may be safely added to other inhibitors of the renin–angiotensin–aldosterone system to provide a greater level of inhibition (O’Brien et al., 2007). Due to its cost and a relative lack of experience in its use, it can only be suggested as an add-on therapy where other more established treatment options have failed to control blood pressure. It is generally well tolerated but may cause diarrhoea at higher doses.

The endothelin antagonist darusentan is undergoing clinical trials in resistant hypertension. Early studies show it may be effective in resistant cases but may be associated with a high incidence of fluid retention (Weber et al., 2009).

Drug selection

Drugs should be chosen on the basis of efficacy, safety, convenience to the patient and cost. For assessing efficacy, it is essential to use evidence from large-scale clinical trials that demonstrate measurable effects on hard end-points like incidence of stroke and other cardiovascular events or death. Smaller scale studies looking at the effects of drugs on blood pressure and surrogate markers such as left ventricular hypertrophy or carotid artery stenosis may generate a hypothesis of a future treatment strategy but should be used to change current strategies. When considering safety, it is important to recognise that these drugs will be taken in the long term and there are advantages to using drugs which have long-established safety records. It is also important to recognise the importance of symptomatic adverse effects since these may reduce adherence. Patients should feel as well during treatment of their blood pressure as they did before drug treatment was instituted. Patient convenience is another important factor and use of once-daily preparations will result in better adherence than more frequent regimens. Since the hypertensive population is very large it is necessary to be conscious of the cost of individual preparations. Combinations of low doses of antihypertensive drugs are often better tolerated than single drugs taken in high dose. The choice of drugs available for treating hypertension is shown in Table 19.3, and common therapeutic problems are noted.

Clinical trial evidence

Initial evidence of benefit in placebo-controlled clinical trials came from studies that primarily involved thiazide diuretics or β-blockers. However, there is increasing evidence of clinical benefit from newer drug classes including ACE inhibitors and calcium channel blockers.

The Blood Pressure Lowering Treatment Trialists Collaboration (2000) carried out a meta-analysis of old against new treatments. They concluded that newer treatments were no more effective than older therapies. Since this study was done several landmark comparative clinical trials have been published.

The Captopril Prevention Project (CAPPP) demonstrated that captopril was as effective as diuretics or β-blockers for preventing cardiovascular morbidity (Hansson et al., 1999a). However, captopril was associated with a 25% higher stroke risk, perhaps because it did not reduce blood pressure as effectively as conventional therapy in this particular study.

The LIFE study demonstrated that losartan was more effective at preventing vascular events, especially stroke, than atenolol in just over 9000 hypertensive patients with left ventricular hypertrophy, although reductions in blood pressure were similar. Losartan was also better tolerated (Dahlöf et al., 2002).

The ALLHAT study (ALLHAT Collaborative Research Group, 2002) involved over 40,000 older, high-risk hypertensive patients with the aim of determining whether the occurrence of fatal coronary heart disease or non-fatal myocardial infarction was lower in those treated with newer agents (amlodipine, lisinopril or doxazosin) compared with a thiazide-like diuretic (chlortalidone). The doxazosin arm was discontinued early because of a higher rate of events, especially heart failure, compared with the diuretic. For the remaining three drugs there was no difference in occurrence of the primary end-point. Chlortalidone was more effective than amlodipine and lisinopril in lowering blood pressure and preventing heart failure and was also marginally more effective than lisinopril in preventing stroke.

The second Australian National Blood Pressure Study Group (Wing et al., 2003) compared enalapril with hydrochlorthiazide in just over 6000 hypertensive subjects recruited in primary care. The primary end-point was any cardiovascular event or death from any cause. In this relatively small study, there was a trend in favour of the ACE inhibitor which was of borderline statistical significance.

The VALUE study (Julius et al., 2004) compared amlodipine and valsartan in high-risk hypertensive subjects. No differences in the primary composite cardiac end-point were observed, although non-fatal myocardial infarction was less common with amlodipine, which also lowered blood pressure to a greater extent. Conversely, onset of diabetes was less common with valsartan.

The ASCOT study (Dahlöf et al., 2005) compared a modern treatment regimen based on amlodipine and perindopril with a traditional regimen based on atenolol and bendroflumethiazide. The study involved over 20,000 high-risk hypertensives. The amlodipine-based therapy was associated with better blood pressure reduction and reductions in the occurrence of cardiovascular events, total mortality and diabetes, although the primary composite end-point was not significantly affected. It is uncertain how much of the benefit can be attributed to the better blood pressure control achieved in the amlodipine-based arm and how specific these findings are to the drug doses and sequencing specified in the trial protocol for each arm of the study.

These various trials have provided results that are conflicting, in part because of differences in trial design and quality. However, there is increasing evidence that β-blockers may be less effective at preventing cardiovascular end-points, as suggested by LIFE and ASCOT studies. In a meta-analysis (Lindholm et al., 2005), β-blockers were less effective than other antihypertensives at preventing stroke, although no significant differences were observed in effects on myocardial infarction or death. There is no consistent evidence that thiazides or thiazide-like drugs are less effective than newer agents in preventing cardiovascular events.

Recommendations for drug sequencing

In the UK, the BHS and NICE issued joint guidelines in 2006 on the order in which drugs should be used (National Collaborating Centre for Chronic Conditions, 2006). This replaced previous guidelines which differed between the two reflecting the ongoing controversy regarding the role of β-blockers. These recommend an initial choice of an ACE inhibitor or ARB (A) as first-line therapy in younger (<55 years) non-black patients. The rationale for this is that these patients often have hypertension associated with high concentrations of renin. It is, therefore, logical to treat these patients with drugs that antagonise the renin–angiotensin system.

For patients >55 and black patients, who tend to have hypertension associated with low renin concentrations, calcium channel blockers (C) or thiazide diuretics (D) are advocated as first-line options. If initial drug therapy fails to control blood pressure, A and D or C is suggested. Subsequently, a combination of A plus C plus D may be used. After this, further therapies, for example, β-blocker, α-blocker, spironolactone, etc., could be added as necessary to achieve adequate control (Fig. 19.1). β-Blockers may be used in those patients with a high sympathetic drive, in pregnant women where labetolol has a good safety record or where other agents are not tolerated.

European guidance (Box 19.3) has eschewed a formal ranking of treatments and instead suggested a table of drugs and indications where they might be most appropriately indicated (Mancia et al., 2007).

Race

People of African Caribbean origin have an increased prevalence of hypertension and left ventricular hypertrophy and are at high risk of stroke and renal failure. They obtain particular benefit from reduced salt intake and are also sensitive to diuretic and calcium channel blockers, while β-blockers appear less effective, at least when used as monotherapy. African Caribbean people have reduced plasma renin activity and, as a result, ACE inhibitors and ARBs are also less effective. This was illustrated in the ALLHAT study where stroke and coronary events were more common in black patients randomised to lisinopril compared to those receiving chlortalidone.

British Asians also have an increased prevalence of hypertension, diabetes and insulin resistance and a particularly high risk of coronary heart disease and stroke. There is currently no evidence of a difference in drug response when compared with white Europeans. However, combinations of β-blockers and thiazides should be avoided when possible because of the higher risk of diabetes.

Elderly

The elderly have a high prevalence of hypertension, with over 70% having blood pressures greater than 140/90 mmHg. They are also at high absolute risk of cardiovascular events. Therefore, the absolute benefits of blood pressure treatment are particularly large in this group. Antihypertensive therapy may also reduce the risk of heart failure and dementia. The Study of Cognition and Prognosis in the Elderly (SCOPE) study (Lithell et al., 2003) was designed to investigate the effects of candesartan on the occurrence of cognitive decline or dementia but revealed no benefit, probably because of the lack of difference in blood pressure between the two arms of the study

The elderly are at particular risk of certain adverse effects of treatment such as postural hypotension and it is important that both sitting and standing blood pressure are monitored. Nevertheless, the benefits of therapy are so great that treatment should be offered at any age unless the patient is very frail or their life expectancy is very short. Isolated systolic hypertension (systolic >160 mm Hg, diastolic <90 mmHg) is common in the elderly and there is irrefutable evidence that drug treatment is beneficial in this group (SHEP Co-operative Research Group, 1991, Staessen et al., 1997). The elderly have more variable blood pressure and larger numbers of measurements may be required to confirm hypertension.

Calcium channel blockers and low-dose thiazide diuretics are safe and effective treatments for elderly hypertensive people and their use is endorsed by large-scale clinical trials. β-Blockers are less effective at reducing blood pressure and preventing clinical end-points. The Swedish Trial in Old Patients with hypertension-2 (STOP-2) compared the effects of conventional (β-blocker or thiazide) and newer drugs (ACE inhibitors or calcium channel blockers) on cardiovascular morbidity in older subjects and did not detect significant differences (Hansson et al., 1999b).

In the Hypertension in the Very Elderly Trial (HYVET), 4000 patients with a mean age of 84, blood pressure 160–199 mmHg systolic at entry were treated to a target of systolic 150 mmHg for 1.8 years with indapamide (a thiazide-like diuretic) and if required the ACE inhibitor perindopril. There was a 30% reduction in fatal and non-fatal stroke, 21% reduction in death from all causes, and fewer adverse events in the actively treated group (Beckett, 2008).

The elderly certainly benefit from treatment of hypertension but the threshold and target for treatment has not been fully elucidated. Most studies in the elderly recruited patients with relatively high baseline targets (>160–190 mmHg systolic) achieving blood pressure on treatment of between 150 and 170 mmHg and only one achieved target blood pressure lower than 140 mmHg and in this study outcome was poorer in the treated group (JATOS Study Group, 2008). There may be little benefit in striving for strict systolic targets beyond 150 mmHg in these patients, particularly if control is being achieved to the detriment of overall patient well-being.

Diabetes

In type 1 diabetes, the presence of hypertension often indicates the presence of diabetic nephropathy. In this group, blood pressure reduction and ACE inhibition slow the rate of decline in renal function. To achieve adequate blood pressure control, combinations of drugs will be needed. Thiazides, β-blockers, calcium channel blockers and α-blockers are all suitable as add-on treatments to ACE inhibitors which should be first-line therapy. Target blood pressure should be <130/80 mmHg or <125/75 mmHg if there is diabetic nephropathy. The evidence supporting this recommendation is, however, limited and obtaining such levels of control in diabetics often impossible.

In type 2 (non-insulin dependent) diabetes, hypertension is particularly common, affecting 70% of people in this group. It is strongly associated with obesity and insulin resistance and control of blood pressure is more important for preventing complications than tight glycaemic control. There is no evidence that one group of drugs is more or less effective than any other. The ADVANCE trial treated diabetics with indapamide and perindopril in addition to prestudy antihypertensive agents. The active group was found to have a further reduction in blood pressure and a significant reduction in adverse renal outcomes (21%) (Patel et al., 2007). It remains a subject of debate whether ACE inhibitors and ARBs have specific renoprotective benefits over and above their effects on blood pressure.

Renal disease

In patients with chronic renal impairment, good blood pressure control slows the progression of renal dysfunction. ACE inhibition reduces the incidence of end-stage renal failure but it is not clear if this is a specific effect or non-specific action as a result of blood pressure lowering. ACE inhibitors also reduce 24-h protein loss and should be used in patients with 24-h protein excretion of >3 g or rapidly progressive renal dysfunction. ACE inhibitors may worsen renal impairment in patients with renal vascular disease and careful monitoring of electrolytes and creatinine is mandatory. Salt restriction is particularly important in managing hypertension in renal disease. Thiazide diuretics are ineffective in patients with significant renal dysfunction and loop diuretics should be used when a diuretic is needed.

A further note of caution regarding overtreatment of blood pressure to overaggressive targets comes from the ONTARGET study which randomised patients with vascular disease or high-risk diabetics to high-dose ramipril, telmisartan (ARB) or both. Many patients were already taking polypharmacy for hypertension and blood pressures at entry were approximately 142/82 mmHg in all groups. Treatment reduced blood pressure by 6.4/4.3 mmHg in the ramipril group, 7.4/5.0 mmHg in the telmisartan group and 9.8/6.3 mmHg in the combination group. This would give the combination group a post-treatment blood pressure of 132/76 mmHg. This combination group was associated with adverse renal outcomes, for example, renal failure and high potassium with no improvement in other cardiovascular outcomes (Yusuf et al., 2008a).

Stroke

Hypertension is the most important risk factor for stroke in patients with or without previous stroke. There is increasing evidence that in those with a previous stroke, blood pressure reduction reduces the risk of stroke recurrence as well as other cardiovascular events. The PROGRESS study, while clearly demonstrating a benefit of lowering blood pressure in patients with cerebrovascular disease, only demonstrated benefit in those whose blood pressure was >140 mmHg on entry or who were already on antihypertensives. The size of benefit was proportional to the size of the blood pressure reduction. The combination of perindopril and indapamide lowered systolic blood pressure by 12.3 mmHg and stroke incidence by an impressive 43%, while perindopril alone was associated with a small drop in systolic blood pressure and no reduction in stroke risk (PROGRESS Collaborative Group, 2001). On treatment, blood pressure in the actively treated group was 132 mmHg systolic which has led some to suggest a target of 130 mmHg for such patients but this is based on post hoc analysis and has not been recommended in formal guidance.

The PROFESS study randomised patients with cerebrovascular disease to telmistartan or placebo and obtained systolic blood pressure of 136 in the active group versus 140 mmHg in the placebo group with no difference in the vascular outcomes between the two groups. This may have been due to the relatively small difference in blood pressure between the two groups (Yusuf et al., 2008b).

The question ‘what to do with blood pressure in the setting of acute stroke?’ has remained an evidence-free zone until fairly recently. Blood pressure naturally rises then falls in the days and hours following acute stroke and some have argued that elevated levels are necessary to maintain brain circulation due to the failure of cerebral autoregulatory mechanisms around the time of stroke. The theory that lowering blood pressure could reduce cerebral perfusion due to a lack of the usual autoregulatory mechanisms is counterweighted by the potential for further damage due to cerebral oedema. The Control Hypertension and Hypotension Immediately Poststroke Study (CHHIPS) randomised acute stroke patients to placebo, lisinopril (sublingual) or intravenous bolus of labetolol and evaluated the incidence of neurological deterioration. There was no adverse outcome in any actively treated group despite reductions in blood pressure (21 vs. 11 mmHg for systolic blood pressure; Potter et al., 2009) This was a relatively small study and larger confirmatory studies are required before firm recommendations for patient management should be made.

In patients with intracerebral haemorrhage, acute reduction of blood pressure has also been demonstrated to be feasible and probably safe with reduced haematoma growth in the actively treated group (Anderson et al., 2008).

Pregnancy

An increased blood pressure before 20 weeks gestation usually indicates pre-existing chronic hypertension that may not have been previously diagnosed. As in all younger hypertensive patients, a careful assessment is needed to exclude possible secondary causes, although radiological and radionuclide investigations should usually be deferred until after pregnancy. Hypertension diagnosed after 20 weeks gestation may also indicate chronic hypertension, which may have been masked during early pregnancy by the fall in blood pressure that occurs at that time. Patients with elevated blood pressure in pregnancy are at increased risk of pre-eclampsia and intrauterine growth retardation. They need frequent checks of their blood pressure, urinalysis and fetal growth. Pre-eclampsia is diagnosed when the blood pressure increases by 30/15 mmHg from measurements obtained in early pregnancy or if the diastolic blood pressure exceeds 110 mmHg and proteinuria is present. There is consensus that blood pressure should be treated with drugs if it exceeds 150–160/100–110 mmHg, although some clinicians use a lower threshold, for example, 140/90 mmHg. Methyldopa is the most suitable drug choice for use in pregnancy because of its long-term safety record. Calcium channel blockers, hydralazine and labetalol are also used. β-Blockers, particularly atenolol, are used less often as they are associated with intrauterine growth retardation. Although diuretics reduce the incidence of pre-eclampsia they are little used in pregnancy because of concerns about decreasing maternal blood volume. ACE inhibitors and ARBs are contraindicated, as they are associated with oligohydramnios, renal failure and intrauterine death.

Meta-analysis of trials suggests that antihypertensive drugs reduce risk of progression to severe hypertension and reduce hospital admissions, although excessive blood pressure reduction may reduce fetal growth.

Oral contraceptives

Use of combined oral contraceptives results, on average, in an increase of 5/3 mmHg in blood pressure. However, severe hypertension can occur in a small proportion of recipients months or years into treatment. Progesterone-only preparations do not cause hypertension so often but are less effective for contraception, especially in younger women. Combined oral contraceptives are not absolutely contraindicated in hypertension unless other risk factors for cardiovascular disease, such as smoking, are present.

Hormone replacement therapy

There is little evidence that hormone replacement therapy is associated with an increase in blood pressure and women with hypertension should not be denied access to these agents if there is an appropriate indication. However, hormone replacement therapy itself does not reduce and may increase the risk of cardiovascular events. Large increases in blood pressure have occasionally been reported in individuals and it is important to monitor blood pressure during the first few weeks of therapy and 6-monthly thereafter. In women with resistant hypertension, during treatment with hormone replacement therapy, the effectiveness of discontinuing hormone replacement should be assessed.

A list of the indications and contraindications to the various antihypertensive agents can be found in Table 19.4.

Table 19.4 Use of antihypertensive drugs adapted from British Hypertension Society guidelines

Note: Strong indications and contraindications are shown. Text in parentheses indicates weak/possible indications or contraindications.

Aspirin

The use of aspirin reduces cardiovascular events at the expense of an increase in gastro-intestinal complications. Its use should be restricted to patients who have no contraindications and either:

Blood pressure should be controlled (<150/90 mmHg) before aspirin is instituted.

Lipid-lowering therapy

There is increasing evidence from clinical trials of the benefit of lipid-lowering drug treatment in patients with hypertension. For example, in the ASCOT study lipid-lowering arm (ASCOT-LLA), treatment with atorvastatin 10 mg was associated with substantial reductions in coronary heart disease and stroke, in spite of the fact that those with total cholesterols initially higher than 6.5 mmol/L were excluded from the study (Sever et al., 2003). Lipid-lowering therapy, usually with a statin, should be prescribed to patients under 80 years of age with a total cholesterol >3.5 mmol/L who either have pre-existing vascular disease or a 10-year cardiovascular risk of >20%.