Some of the studies have included internal comparison groups, which allow a better estimation of the association between acromegaly and hypertension. One group compared the frequency of hypertension between a retrospective series of 158 acromegalic subjects and a large, local epidemiological study [23], only finding a raised prevalence of hypertension in female patients. Hypertension was found in 19.6% of acromegalics. However, when the figures were recalculated using modern definitions, hypertension is considerably more common in both sexes, with 50.6% of patients being hypertensive vs. 32.9% in the comparison population. A Japanese group found that the prevalence of hypertension was higher in a retrospective cohort of 64 acromegalics (37.5%) than in the local population (20–25%) [25]. They found that 75% of patients with acromegaly classified as hypertensive (using the old definition) had a family history of hypertension compared to 25.9% of normotensive acromegalics. This effect may have been diluted when hypertension is defined by modern standards as only 23.1% of their “borderline hypertensive” acromegalics had a positive family history. However, since a family history of hypertension increases the risk of hypertension in non-acromegalics as well, this may be a confounder rather than a specific effect.

A larger subsequent study used 200 cases and controls matched for age, sex, body mass index, and lifestyle factors and found hypertension in 46% of acromegalic cases and 25% (p  <  0.0001) of controls [14]; cases had significantly higher diastolic and systolic BPs. In contrast to some other studies, gender did not influence the acromegaly-hypertension association; they also found that a family history of hypertension was more common in controls than patients, in contrast with Kraatz et al. [23]. Hypertension was positively correlated with age in both groups, although it occurred 10 years earlier in cases than in controls. The strength of this study lies in its relatively large size and the use of controls matched for factors that could otherwise confound the results (the prevalence also agrees with worldwide estimates [16]); however, there is no mention of blinding of investigators which may be a weakness.

One study looked at hypertension in a cohort of 57 patients who were diagnosed with acromegaly when older than 60, comparing them with gender and age-matched controls [26]. They found that 82% of the patients, whereas only 67% of controls were hypertensive. Mean systolic and diastolic BPs were significantly higher in cases. The authors note that many of the acromegalic patients are being treated unsuccessfully for hypertension. This may imply that hypertension associated with acromegaly is relatively refractory to treatment.

Two studies have compared controls with a mix of acromegalic patients—treatment naïve, and those receiving or who recently received unsuccessful treatment of their acromegaly. Sucunza et al. found that 37.7% of acromegalic patients (n  =  19) were hypertensive compared with only 15% of controls (n  =  105) [27]. Ciresi’s group also did a small study of acromegaly patients mostly receiving therapy, 12 of whom were not responding completely. They compared the 12 with active acromegaly against 21 controls and found hypertension in 66.7% of cases and 0% of controls [15]. These papers do corroborate the evidence from other stronger studies; however, they are both small and contained many patients receiving acromegaly treatment, a potentially confounding attribute as a search of the literature shows that somatostatin analogues (SSA), a mainstay of medical therapy for acromegaly, may themselves increase BP [28].

Thus, despite the wide range of prevalence findings, the association between acromegaly and hypertension appears to have a strong base of evidence. The best evidence comes from the Naples group’s case–control study of 400 subjects [14]; their result that 46% of patients with acromegaly are hypertensive by clinical assessment (compared with 25% of controls) also falls around the mean of the other studies’ results. This does not, however, provide evidence of a causal relationship.

One observation that would provide a stronger argument for causality is if treatment of acromegaly reduces or resolves hypertension. The evidence surrounding this is mixed, with some studies finding a reduction in BP while others showing no effect. There may also be differences between the different methods of treatment—medical vs. surgical.

The best study to address this question may be that of Colao et al. from 2008 [13]. They retrospectively selected 105 treatment-naïve acromegalic patients, all receiving therapy for hypertension. Hypertension was successfully controlled in only 14% of patients; the remainder continue to be hypertensive despite therapy. Subjects received a variety of treatments, which normalized GH and IGF-I levels in 72%. In this group both systolic and diastolic mean BP fell by 13.1% and 11.4% respectively (p  <  0.0001), independent of treatment method. The reduction in BP was not statistically significant in patients whose acromegaly therapy was unsuccessful. This study of hypertensives found that basal systolic and diastolic BPs were correlated with GH and IGF-I levels. Furthermore, there was a significant correlation between the percentage reduction in GH/IGF-I levels and the percentage reduction in systolic and diastolic BPs.

Their findings of both a degree of dose–response relationship between GH/IGF-I levels and BP, and a significant reduction in BP and uncontrolled hypertension when acromegaly is controlled represent the strongest evidence yet that acromegaly does cause hypertension. Limitations of the study include the exclusion of hypertensive patients who were not receiving antihypertensives and the lack of blinding. These may have biased results, however, the overwhelming strengths of this study designed—unlike others—specifically to detect a change in BP with acromegaly treatment lend considerable weight to the argument that acromegaly causes a potentially reversible hypertension. The subsequent year the same group produced a study looking at patients (some of whom have also been included in the previous study after shorter follow-up) who received only SSA treatment for 5 years (excluding patients who were initially or subsequently managed surgically) [29]. They observed a reduction in hypertension greater than 50% (p  =  0.027). This reduction retained significance when patients were analyzed by sex or type of SSA.

One study of 31 patients (11 hypertensives) who underwent successful surgery for acromegaly measured pre- and post-operative BP clinically and using ABPM. They found significant reductions in systolic and diastolic BPs by clinical measurement [20]. With ABPM they found that 24 h, daytime and nighttime systolic pressures fell significantly; however, corresponding diastolic changes were nonsignificant.

These papers provide focused data about whether treatment of acromegaly improves the associated hypertension by removing confounding diluent effects in other papers of small initial numbers of hypertensive patients and patients who do not respond adequately to treatment. Another paper by Delaroudis and colleagues [30] found a significant reduction in BP in 18 patients on SSAs that achieved only partial control of their GH/IGF-I levels. A Russian group recently presented data on a sample of 232 acromegalic patients, 80.2% of whom were hypertensive. They report a normalization of BP in patients whose acromegaly began before age 45, but not in others [31]. Other groups have found a link between age and cardiovascular comorbidities in acromegaly. A phenomenon also recognized in acromegalic cardiomyopathy [32], duration of disease, is likely to play a role in this. A Dutch study looking at comorbidities of patients in long-term remission from acromegaly found that compared to epidemiological data, only patients aged 50–70 were significantly more hypertensive than controls [33]. Some groups looking at whether remission of acromegaly results in resolution of hypertension have found less or no association. It should be mentioned that few of these studies focus on hypertension as a primary outcome; most are looking at changes in other conditions, usually cardiomyopathy.

Colao et al. published another study in 2008 looking at improvement in cardiovascular morbidity by different successful acromegaly treatments: surgery vs. SSAs [34]. In this case only diastolic BP decreased significantly, but with both treatment approaches. Overall, there was a nonsignificant trend to a reduction in the number of hypertensives with both treatments. This may not have reached statistical significance because the analysis subdivided hypertension into stages 1 and 2 and patients were divided by treatment, meaning small samples. A significant reduction in diastolic BP alone was also found in another paper when GH/IGF-I levels were normalized [35]. Another study found that patients in remission after surgical treatment of acromegaly had significantly lower diastolic BP than those whose surgery had failed, but hypertension prevalence and systolic BP were not significantly different [36].

A 2007 meta-analysis of 18 papers (none were randomized control trials) which examined the effect of SSAs on the heart in acromegaly found no significant effect on hypertension [37]. However, hypertension was not one of the inclusion criteria, meaning certain papers were ignored including more recent studies that used the best designs for detecting changes in hypertension. Another reason that the meta-analysis may have failed to detect a significant difference is that many of the studies were done before the currently accepted definition of remission of acromegaly. The mean level of post-treatment GH after an oral glucose tolerance test (OGTT) was 7.6 μg/L—today an accepted level for remission is <1 μg/L [38]. This incomplete remission of acromegaly may have meant that changes in BP were insignificant; however, another group have found significant changes with partial control of ­acromegaly (although mean nadir GH was lower at 2.35 μg/L) [30]. Other potential explanations include that the possibility that follow-up was not sufficiently long to see resolution of hypertension, possible effects of different age demographics, that heterogeneity influenced results or that low numbers of hypertensives meant that mean BP changes were small when diluted with non-hypertensive patients.

Apart from this meta-analysis, papers from other groups found no significant reduction in BP or hypertension. One series including 15 patients (4 hypertensives) [39] found no significant reduction. Two reasonably sized (43 and 61 patient) studies compared those with successful remission of acromegaly with those with partial remission and with controls; neither found any difference in BP or hypertension [40, 41]. One study of 10 patients treated with a subcutaneous SSA (octreotide) found a significant increase in mean 24 h and nighttime systolic and diastolic BPs with ABPM [28]. They hypothesize that this may be due either to vasoconstrictive effects of octreotide on the splanchnic vasculature, increasing systemic circulating volume, or because ABPM provides a better assessment of hypertension than clinical measurements. It seems possible that SSAs produce this effect only transiently, and if other studies measure BP clinically at another time to treatment this effect might be missed. Perhaps only ABPM, with its long-term measurements would detect this change.

The combination of these findings presents a mixed picture. Methodologically, the strongest studies do find a reduction in BP. However, we cannot ignore evidence from the meta-analysis and other studies that found no change. Despite having flaws in their design for detecting change in BP (most not having been designed with this aim) they demonstrate that BP reduction with treatment of acromegaly may only happen in certain patients or settings, with certain treatments, or that changes are small. More studies are required, including the use of ABPM and larger cohorts looking for changes in BP in both hypertensives and normotensives. Samples need to be large so that data can be analyzed for those whose treatment is successful and for those who remain acromegalic.

A very small number of groups have attempted to elucidate the nature of the BP change that occurs in acromegaly. As seen above some have found that hypertension is predominantly diastolic or systolic, however there is also evidence from studies using ABPM about BP patterns in acromegaly, particularly loss of the normal circadian fall in nocturnal BP or “non-dipping.” Non-dipping has been found to be independently associated with significant cardiovascular morbidity [42]; two population studies have estimated its prevalence in a cohort of Swedish men (6.2%) [43] and in normotensive non-diabetic Finnish subjects (18.8%) [44]. Jaffrain-Rai and colleagues found that 44.1% of 68 acromegalic patients were non-dippers [18]. They analyzed to look for interaction with the glucose tolerance of patients—since diabetes mellitus is independently associated with non-dipping—but no such interaction existed. They do highlight the important fact that lack of nocturnal fall in BP may be confounded by the coexistence of sleep apnea—this could cause nocturnal hypertension and is found in acromegaly. Another study (n  =  25) found a high prevalence of non-dipping in both normotensive patients (40%) and hypertensive patients with acromegaly (60%), and that non-dipping was associated with higher GH levels [19]. These agreed with a previous smaller paper [21] but contradict findings of Minniti et al. who found only one of seven hypertensive acromegalic patients were non-dippers [17]. The latter paper also found a very large difference between the prevalence of hypertension when using clinical measurements or ABPM (as mentioned above) so the population may have been atypical. Another group have presented data showing similarly high levels of non-dipping (57%) in acromegalics regardless of whether hypertensive or not [45]. The only evidence where controls were used is from a small study that found non-dipping was more common in acromegalics (62%) vs. controls (12%, p  <  0.01) [21]. In summary, it appears that the non-dipping profile may be associated with acromegaly, beyond that expected for the hypertension and impaired glucose tolerance seen in the condition. Whether this effect is mediated by concomitant obstructive sleep apnea remains to be clarified.

Evidence from case series and case–control studies suggest strongly that acromegaly is associated with elevated BP. The prevalence remains unclear, with estimates ranging from 17.5% to over 80%, however most studies, particularly larger ones designed to look at hypertension in acromegaly find it to be between approximately 40 and 60%, and higher than in matched controls. Acromegaly may also cause the non-dipping nocturnal BP pattern.

As mentioned above, hypertension is of significance in acromegaly because of its negative impact on acromegalic cardiomyopathy and as a predictor of mortality. It appears that treatment of acromegaly can lead to an improvement of hypertension; however, there is not sufficient evidence about effects of management of hypertension on mortality in acromegaly. It also remains to be clarified whether the effects are directly from GH/IGF-I excess or indirectly e.g. from obstructive sleep apnea; there does not seem to be a consistent correlation between either GH/IGF-I levels or duration of disease and degree of hypertension [20], pointing to a complex association.