Obese (defined as a BMI greater than 30 kg/m2) adolescents with hypertension experience a marked fall in BP after weight-loss following bariatric surgery, with 74 % becoming normotensive (Inge et al. 2016). In younger subjects, obesity (particularly central) is linked to a significant increase in sympathetic nerve activity (Grassi et al. 2004). The obesity-related high sympathetic nerve activity may be confined to men (Kostis et al. 2015), and be apparent mainly in muscle and kidney (Brooks et al. 2015). Obesity-related increases in sympathetic nerve activity are particularly apparent in the presence of hypertension (Lambert et al. 2007)- Fig. 1, or type-2 diabetes (Huggett et al. 2003). Even high-normal blood pressure is linked to increased muscle sympathetic nerve activity (Seravalle et al. 2015). The raised sympathetic nerve activity is associated with the release of leptin (so-called “thin hormone”) from adipose tissue; leptin acts upon the hypothalamic region of the mid-brain, resulting in increased sympathetic nerve activity (Barnes and McDougal 2014). High insulin levels, associated with obesity-related insulin resistance, also act upon the hypothalamic region, resulting in heightened sympathetic nerve activity (Coats and Cruickshank 2014; Cruickshank 2014). This has important prognostic implications, as high norepinephrine (noradrenaline) levels are associated with the atherosclerotic process (Helin et al. 1970) and (via an increased heart rate) coronary plaque rupture (Heidland and Strauer 2001a). It thus comes as no surprise that high plasma norepinephrine levels, independent of smoking and blood pressure levels, are powerful predictors of cardiovascular death and survival in 601 (354 men and 247 women) young/middle-aged hypertensive subjects (over a 6–7 year follow-up period (Peng et al. 2006) – Fig. 2. Importantly, high intra-lymphocyte beta-receptor density (Bmax) and cyclic adenosine monophosphate (AMP) levels predict (independent of BP) future myocardial infarctions, but not stroke (which relates to blood pressure) (Peng et al. 2006) – Fig. 3. In the Framingham Heart Study (Gillman et al. 1993), high resting heart rates, particularly over 85 bpm (a surrogate for increased sympathetic nerve activity), in young/middle-aged hypertensive subjects, have been shown to predict all-cause death and cardiovascular and coronary heart disease events for both hypertensive men – Fig. 4, and women over a 36 year follow-up period.

Though sympathetic nerve activity is increased in elderly hypertension (Yamada et al. 1989; Hart and Charkoudian 2014), this is not so within the kidney (Esler et al. 1984). There is also a marked reduction in beta-receptor affinity/sensitivity in this older age-group (Feldman et al. 1984; Tenero et al. 1990), which may explain the relative lack of efficacy of beta-blockers in the elderly – (see later).

In the above context, it is notable that antihypertensive agents that increase sympathetic nerve activity in young/middle-age, perform poorly in terms of reducing cardiovascular events in this age-group. Thus, thiazide-type diuretics increase sympathetic nerve activity (Menon et al. 2009), and in 3 studies involving diuretic therapy in young/middle-aged hypertensive subjects (No author listed 1980; Medical Research Working Party 1985; Leren and Heigeland 1986) there was no reduction in the risk of myocardial infarction (No author listed 1980; Medical Research Working Party 1985), and even a significant increase (Leren and Heigeland 1986), versus randomised placebo/non-treatment. Dihydropridine calcium blockers (felodipine, amlodipine, manidipine, and lacidipine) increase heart rate and plasma norepinephrine levels (Fogari et al. 2000), and in the ABCD study (Estacio et al. 1998) the investigation was terminated prematurely due to a significant excess of myocardial infarctions in the nisoldipine, vs the enalapril, group. Likewise, angiotensin receptor blockers (ARBs) increase sympathetic nerve activity in younger subjects (Heuser et al. 2003; Moltzer et al. 2010). Meta-analyses indicates that, in contrast to ACE-inhibitors, ARBs increase the risk of myocardial infarction (Strauss and Hall 2006; Straus and Hall 2007) Fig. 5, and in two subsequent placebo-controlled studies involving hypertension (Imai et al. 2011) and pre-hypertension plus diabetes (Haller et al. 2011), there was a significant excess if cardiovascular events in those receiving the ARB. Thus, prevention of myocardial infarction and cardiovascular events is not just about good control of BP. In contrast to ARBs, ACE-inhibition results in a reduction in sympathetic nerve activity (Noll et al. 1997) – see later.

As myocardial infarction is the most common cardiovascular event in young/middle-aged essential hypertension (Medical Research Working Party 1985), it is important to note that beta-blockade is able to reverse the coronary atherosclerotic process. In a pooled analysis of four intravascular ultrasonography randomised, controlled trials in patients with coronary heart disease, over an 18–24 month time-interval, BBs (mainly atenolol and metoprolol) effected a significant (p < 0.001) regression of coronary artery atherosclerotic plaque (Sipahi et al. 2007). BBs also stabilise vulnerable coronary plaque. A study of 106 middle-aged patients who underwent coronary angiography twice within a 6 month period, revealed that high heart rates (>80 bpm) significantly increased, and BBs significantly decreased, the risk of atheromataous plaque disruption and rupture (a precursor to myocardial infarction) (Heidland and Strauer 2001b).

Beta-2 blockade results in a rise in BP of about 7/5 mm Hg (Robb et al. 1985). Thus a moderately beta-1 selective agent like atenolol is more effective in lowering BP than a non-selective BB like propranolol (Zacharias and Cowen 1977). Atenolol, in turn, is less effective in lowering BP than highly beta-1 selective bisoprolol (Neutel et al. 1993). Indeed, in younger/middle-aged hypertensive subjects, bisoprolol is a more effective anti-hypertensive agent than the calcium blocker amlodipine, the alpha-blocker doxazosine, the ACE-inhibitor lisinopril, and the diuretic bendrofluozide (Deary et al. 2002), and angiotensin receptor blockers (ARBs) (Hiltunen et al. 2007), being at least as reno-protective as the latter (Parrinello et al. 2009).

As already noted, recent meta-analyses that do not take age into account, are less than complimentary to BBs (Wu et al. 2013; Prospective Studies Collaboration 2002; Wiysonge and Opie 2013; Lindholm et al. 2005; Xue et al. 2015; Thomopoulos et al. 2015a, b; Dahlof et al. 2002). One (Wu et al. 2013) suggested that BBs increase all-cause mortality; another (Lindholm et al. 2005) indicated that first-line beta-blockade did not reduce all-cause mortality and was associated with only modest reductions in cardiovascular events (vs randomised placebo or non-treatment); another (Xue et al. 2015) suggested that BBs increase the risk of stroke; another (Thomopoulos et al. 2015a) indicated that BBs were inferior to renin-angiotensin system (RAS) inhibitors in preventing cardiovascular events and stroke; and finally 2 meta-analyses (Thomopoulos et al. 2015b; Dahlof et al. 2002) concluded that BBs were less effective than other agents in preventing stroke, and that reduction in coronary heart disease and all-cause death did not achieve statistical significance.

Two meta-analyses that do take age into account (Khan and McAlister 2006; Kuyper and Kahn 2014) arrive at very different conclusions to meta-analyses that do not. One meta-analysis (Khan and McAlister 2006) included 8 randomised placebo-controlled studies, and 9 randomised studies involving active agents (Williams 2007; Medical Research Working Party 1985; The IPPPSH Collaborative Group 1985; Coope and Warrender 1986; Dahlof et al. 1991; MRC Working Party 1992; Trial of secondary prevention with atenolol after transient ischemic attack or nondisabling stroke. The Dutch TIA Trial Study Group. Stroke 24:543–548 1993; Eriksson et al. 1995; Wilhelmsen et al. 1987; UK Prospective Diabetes Study Group 1998; Hansson et al. 1999a, b, 2000; Zanchetti et al. 2002; Pepine et al. 2003; Black et al. 2003; Dahlof et al. 2005). Compared to randomised placebo, in the younger/middle-aged hypertensive subject, with a mean age less than 60 years old, (admittedly and arbitrary cut-off point re definition of young/middle-aged) BBs were significantly superior to placebo in reducing the risk of death/stroke/MI – Fig. 6; with only a positive trend in the elderly – Fig. 7. When compared to randomised comparator antihypertensive drugs, there was a trend favouring BBs in younger hypertensive subjects Fig. 8, in contrast to those older than 60 years old, where BBs were significantly less effective in reducing the risk of death/stroke/myocardial infarction –Fig. 9. The other meta-analysis (Kuyper and Kahn 2014) involved 21 studies, comprising the same 17 studies as the first meta-analysis (Khan and McAlister 2006), plus an extra 4 studies in the younger group (3 studies compared propranolol with diuretic therapy (Veterans Administration Cooperative Study Group on Antihypertensive Agents 1982; Berglund et al. 1986; Yurenev et al. 1992), and the fourth – AASK study (Wright et al. 2002), compared metoprolol with amlodipine and ramipril in Black American hypertensive patients with renal disease). The conclusion was that in the young/middle-aged (less than 60 years old) both atenolol and non-atenolol beta-blockers were similarly effective in reducing cardiovascular endpoints, while in the elderly, atenolol (no other BBs have been studied) was associated with an increased risk of stroke. The second meta-analysis did not include the most recent results of the AASK study (Norris et al. 2006) in younger/middle-age subjects, which showed that metoprolol, amlodipine, and ramipril were similarly effective in reducing cardiovascular outcomes after 4 years of follow-up.

BBs thus have no role to play as first-line agents in the elderly hypertensive subject, unless myocardial ischemia is also present (Pepine et al. 2003), where atenolol was equivalent to the calcium blocker verapamil. The role of BBs in the elderly is as second-line therapy to either diuretics or calcium blockers, as evidenced in the MRC-elderly study (MRC Working Party 1992) and the large ALLHAT (The ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group 2002) and SHEP (SHEP Cooperative Research Group 1991) studies, especially in the presence of the metabolic syndrome (SHEP Cooperative Research Group 1991).

There is a perception that BBs do not reduce all-cause mortality (Wu et al. 2013; Prospective Studies Collaboration 2002; Wiysonge and Opie 2013; Lindholm et al. 2005), and are relatively ineffective in reducing the risk of stroke (Krause et al. 2011; Xue et al. 2015; Thomopoulos et al. 2015a, b; Dahlof et al. 2002; Williams 2007; Khan et al. 2009 May), in younger/middle-aged hypertensive subjects. The UKPDS-39 (UK Prospective Diabetes Study Group 1998) and MRC-1 (Medical Research Working Party 1985) studies give no credence to these perceptions. Concerning stroke-risk reduction in MRC-1 (Medical Research Working Party 1985) in non-smoking men, both non-selective propranolol and the diuretic bendrofluazide, vs randomised placebo, reduced the risk of stroke by 54 %. In UKPDS-38 (U.K. Prospective Diabetes Study Group 1998), where smoking was not taken into account, tight control (either atenolol or captopril) of blood pressure, vs less tight control (difference 10/5 mm hg), resulted in a significant 44 % reduction in the risk of stroke. UKPDS-39 (UK Prospective Diabetes Study Group 1998) examined the effect of atenolol and captopril in reducing macrovascular and microvascular complications over a 9 year follow-up period. Figure 10 shows the effect of the 2 agents in reducing the 7 primary endpoints (plus heart failure- a secondary end point) vs less-tight control of blood pressure. It is apparent that all 8 trends favoured the BB (over the ACE-I) which reduced stroke-risk by about 50 %, peripheral arterial disease-related endpoints by about 60 %, microvascular (kidney and eye) endpoints by about 45 %, and heart failure by about 65 %. Thus, the UKPDS-39 (UK Prospective Diabetes Study Group 1998) results, like MRC-1 (Medical Research Working Party 1985), deny totally the claim that BBs are relatively ineffective in preventing stroke in young/middle-aged hypertensive subjects.

The UKPDS patients were monitored for a further 10 years, with a median total follow-up time of 14.5 years (Holman et al. 2008). The trends favouring the beta-blocker over the ACE-inhibitor tended to persist, but now, in the case of all-cause death, there was a significant 23 % reduction in favour of atenolol – Fig. 11. There is thus no truth in the claim that BBs do not reduce all cause death (Wu et al. 2013; Prospective Studies Collaboration 2002; Wiysonge and Opie 2013; Lindholm et al. 2005).

In middle-aged patients with pre/mild hypertension plus stable myocardial ischemia (von Armin and for the TIBBS Investigators 1996), randomised to either highly beta-1 selective (cardioselective) bisoprolol or nifedipine SR, at 1 year follow-up, event-free survival was significantly superior in those randomised to bisoprolol.

In UKPDS-39, apart from all-cause death, all other benefits of atenolol (relative to ACE-I) noted after the initial study (UK Prospective Diabetes Study Group 1998), tended to diminish with time (mean follow-up time 14.5 years) (Holman et al. 2008). So what other factors might be in evidence?

BBs have been observed by several authors, to modify the initiation and spread of various cancers. Certainly stress has been noted to hasten cancer-progression, probably via activation of tumour-associated beta-receptors by epinephrine and norepinephrine (Cole and Sood 2012; Fitzgerald 2012). Thus, beta-blockade has been noted to benefit 1. Breast cancer and prevent metastases (Cakir et al. 2002; Barron et al. 2011; Melhem-Bertrandt et al. 2011) 2. Colon cancer (Takezaki et al. 2001; Perrone et al. 2008) 3. Pancreatic cancer (Weddle et al. 2001; Zhang et al. 2010) 4. Melanoma (De Giorgi et al. 2012) 5. Lung cancer (Al-Wadei et al. 2012) 6. Neuroblastoma (Pasquier et al. 2013) 7. Prostate cancer (Perron et al. 2004; Grytli et al. 2013) –Table 2. This likely anti-cancer property of BBs is particularly important in the context of the increased risk of cancer in middle-aged hypertensive subjects (Harding et al. 2016).

In three major prospective, randomised, hard-endpoint studies in middle-aged hypertensive subjects, cigarette smoking played a vital role in modifying the potential of the BB to reduce the risk of a cardiovascular event. The MRC-1 study (Medical Research Working Party 1985) compared non-selective propranolol with a thiazide diuretic and placebo, in 17,354 subjects, of whom 30 % of men and 25 % of women were smokers, over a 5 year follow-up period. The IPPPSH study (The IPPPSH Collaborative Group 1985) compared non-selective oxprenolol with placebo, in 6357 subjects, of whom 29 % were smokers, over a 3–5 year follow-up period. The MAPPY study (Wikstrand et al. 1991) (an extension of the HAPPHY study (Wilhelmsen et al. 1987)) compared moderately selective metoprolol with a thiazide diuretic, in 3156 subjects, of whom 34 % were smokers, followed-up for 4 years.

In the case of myocardial infarction (about 3 times more common than stroke in the young/middle-aged hypertensive subject (Medical Research Working Party 1985)), the ability of the BB to reduce the risk of an event by 33–49 % (vs randomised placebo or diuretic therapy) in non-smokers, was not observed in smokers (Medical Research Working Party 1985; The IPPPSH Collaborative Group 1985; Wikstrand et al. 1991). Indeed, in the case of non-selective propranolol and oxprenolol, the risk of myocardial infarction was actually increased by 13–35 % in smokers – Fig. 12. A similar result relating to stroke was also noted in MRC-I (Medical Research Working Party 1985). In the MRC-elderly study (MRC Working Party 1992), atenolol (vs randomised placebo) increased the rate of cardiovascular events by 38 % in smokers, compared to a modest 16 % reduction in non-smokers.