Do antihypertensive medicines increase the risk of depression?

Abstract

Depression is a common and often debilitating mental health condition that affects a significant proportion of the population. Antihypertensive drugs are prescribed for the pharmacological management of high blood pressure, which affects approximately one third of adults globally. While these drugs are generally well tolerated, concerns remain about the potential risk of depression associated with their use. Several studies have investigated the link between antihypertensive drugs and depression, with mixed results. Some studies have reported an increased risk of depression in patients taking certain classes of antihypertensive drugs, such as beta-blockers and calcium channel blockers; others have found no significant association between these medicines and depression, or report a reduction in risk of depression. Using evidence, this review considers whether any differences in the risk of depression exist between antihypertensive classes or if any particular drugs stand out as being more or less likely to be linked with depression. We found no substantial evidence of any class effect on risk of depression, although lipophilic beta-blockers are associated with increased reports of sleep disturbance and fatigue. Given the high prevalence of hypertension and depression, further well-designed studies are needed to clarify the potential link between antihypertensive drugs and depression.

Key words: Antihypertensive, depression, risk/benefit, hypertension

Introduction

For more than 50 years, a link has been made between the use of antihypertensive drugs and the new onset or re-emergence of symptoms of major depression. Since propranolol became available in the 1960s, both anecdotal and, later, more empirical evidence has emerged that antihypertensive drugs of all classes may have some influence on mood disorders and possibly even on the risk of suicide​[1–5]​. The direction of the evidence (increased vs. decreased risk of depression) and expert opinion is mixed, which does not provide much in the way of helpful guidance for prescribers. The majority of antihypertensive prescriptions today are based on guidance from the National Institute for Health and Care Excellence on hypertension​[6]​. This promotes the use of drugs that target the renin–angiotensin system (RAS) — angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) — dihydropyridine calcium channel blockers and thiazide diuretics at the earlier steps of hypertension management, because of their relatively good profiles in reducing cardiovascular events such as stroke and myocardial infarction​[6]​. Other antihypertensive agents, such as beta-blockers and alpha-blockers, have a less favourable risk/benefit ratio and so appear only as adjuncts or where other agents are contraindicated. Other antihypertensive agents, including centrally-acting agents (such as clonidine and α-methyldopa) and vasopressors (such as hydralazine and sodium nitroprusside) are only used in more specialist cases. Nevertheless, there is evidence linking all classes of antihypertensive drugs with depression and other psychological conditions in the clinical literature, with some classes, and especially some agents, making more frequent appearances than others​[7–9]​. This review explores the association and recent evidence, which suggests that the risk of depression — if or where it exists at all — may differ between sub-classes of antihypertensives.

Antihypertensives and depression: a prescribing dilemma

In the late 1960s, just a few years after the first prototype beta-blocker propranolol became available as an antihypertensive drug, the BMJ reported observations of depression in patients taking the drug​[1]​. Even earlier antihypertensives, such as reserpine and α-methyldopa, have also been linked to the development of depressive symptoms​[10,11]​. Observations of adverse effects reported in reserpine users, in particular, formed the basis of the development of the monoamine hypothesis of depression in the 1950s​[12]​. The monoamine hypothesis posits that depression is associated with a deficit of the monoamines serotonin, noradrenaline and dopamine in the central nervous system (CNS), although today it is recognised that this theory likely explains only a part of the underlying pathology of depression. As the armoury of antihypertensive drugs has increased, giving us the wide range of options available in today’s formularies, the question of whether this now very diverse class of drugs influences mood on either a subclinical or clinical level continues to be posed. Over the years, further evidence has emerged supporting arguments that at least some antihypertensive drugs may actually reduce the risk of depression, while others continue to pose an increased risk, or that their influence as a broad class is negligible​[7,8,13–16]​. Two large-scale, prominent trials considered the effect of antihypertensive use on depression and other patient outcomes. The ACCORD (Action to Control Cardiovascular Risk in Diabetes) trial, published in 2012, investigated the effects of intensive versus standard blood pressure control on depression and health-related quality of life in patients with type 2 diabetes mellitus, and found there was no clinically meaningful difference between the two approaches to managing hypertension​[17]​. In 2017, the findings of SPRINT (Systolic Blood Pressure Intervention Trial) were published, which focused on the effect of intensive blood pressure treatment on patient-reported outcomes, concluding that there was no effect of intensive blood pressure reduction on depression​[18]​.

To deliver and receive the best possible person-centred interventions, prescribers and patients will no doubt take an interest in identifying an antihypertensive drug that best meets the patient’s overall needs, including consideration of the adverse effect profile. This may mean that what is recommended for one individual, based on previous experiences or their ability to tolerate certain adverse effects, will not suit another. History of mental health, or risk of depression, should be considered if there is any risk [of mental health conditions/depression] from the antihypertensive drug. However, the question remains: is there a link betweene antihypertensives and depression and, if so, what weight should be applied to this variable when selecting a suitable antihypertensive drug?

Is there a biological basis?

Largely based on observations of changes in mood in patients who were being treated with reserpine to manage their hypertension, the monoamine theory, introduced above, began to emerge in the 1950s and 1960s​[12]​. Reserpine is one of many alkaloids isolated from extracts of the plant Rauvolfia serpentina, and its mode of action as an antihypertensive is understood to be through irreversible blockade of the vesicular monoamine transporters VMAT1 (in neuroendocrine cells) and VMAT2 (in neurons), resulting in the depletion of both central and peripheral sympathetic amines​[19]​. In the 1960s, seminal work by Joseph Schildkraut showed that monoamine depletion, especially of noradrenaline, appeared to be associated with depression​[12]​. Based on the reported observations of several clinicians that patients taking reserpine were more likely to experience symptoms of depression, and taking into consideration reserpine’s presumed mode of action, it was extrapolated that reserpine use was biologically connected with mood disorders, which contributed to the generation of the monoamine hypothesis of depression. However, reserpine is a drug that is associated with a variety of perceived adverse effects and was not well tolerated by many patients at the doses prescribed at the time​[20]​.

Shortly after propranolol became available as an important new development in the management of hypertension, evidence slowly began to emerge that also connected the drug with symptoms of depression​[1,21]​. Propranolol is a non-selective beta-adrenergic antagonist, more commonly known as a beta-blocker, which has lipophilic properties allowing passage across the blood–brain barrier, resulting in both central and peripheral effects​[22]​. Its pharmacology and pharmacodynamic properties are more complex than the class name suggests, but of particular relevance to this review are its weak effects on noradrenaline availability (increased) in the synapses of the CNS and weak blockade of several serotonin receptors, which may impact in some way on arousal and mood​[23,24]​.

The advancement of antihypertensive therapy in the intervening years has seen the rise of numerous other beta-blockers, diuretics, calcium channel blockers and drugs that target the RAS appearing in the British National Formulary and clinical guidelines, but the question of their impact on mood has not been resolved. In contrast with the attempts outlined above to understand the involvement of monoamines, many of the antihypertensives used clinically today are reported to either increase or decrease the risk of depression based, perhaps, on other biological mechanisms​[2]​. Theories have been posed based on genetic differences and neuroinflammatory mechanisms, which remain inconclusive but nevertheless plausible​[25–28]​.

There is mixed evidence for an association between cerebral blood flow and mood in older people. Saper argues that maintaining a systolic blood pressure at or below 120mmHg, as recommended following the results of SPRINT, is insufficient for adequate cerebral perfusion in older people​[29]​. In an epidemiological study, Hildrum et al. found that there is an association between low blood pressure and mood, but a randomised controlled trial by Moonen et al. found no evidence that discontinuing antihypertensives in older people led to any improvement in cognitive function​[30,31]​

A review of the evidence

For this review, 21 articles (see Figure) were identified as exemplars of the range of evidence and opinion available for practitioners to inform their recommendations and shared decision making. Articles were included based on guiding criteria, which were: 1. Articles must be published in English; 2. Articles must report investigation of a possible relationship between use of one or more antihypertensive drugs in humans and any symptoms commonly associated with depression as defined by ICD-10 criteria, either as part of a trial or in routine clinical practice; 3. Studies of any kind reporting primary data or metadata, but not review or opinion articles. Of the 21 studies, 6 explored the association between beta-blockers and depression (including symptoms, formal diagnoses, associated mental health conditions or risk of suicide) and 15 explored the same relationship for 2 or more antihypertensive drugs. The studies were a mixture of systematic reviews of randomised controlled trials with meta-analyses, cohort and case-control studies, cross-sectional and other observational studies. Ten studies concluded that there was no substantial evidence of any association between antihypertensive use, of any class, and depression. Two studies concluded that all of the antihypertensive drugs investigated had a protective effect against depression. One study found dichotomous class effects with drugs divided into “increases risk” and “decreases risk” categories. The remaining 8 studies concluded that there may be an increased risk of depression for one or more of the antihypertensives investigated in the study. These findings are summarised in the Figure​[5,7–9,13–16,32–44]​.

In total, 18 of the studies investigated the association between beta-blockers and depression. Eight of these studies reported an increased risk of new onset or recurrence of a previous diagnosis of depression with beta-blocker use. One study — a cross-sectional study of 3,218 new beta-blocker users — found that the proportion of propranolol users receiving antidepressant prescriptions was 9.5%, several times greater than for any other lipophilic or hydrophilic beta-blocker​[43]​. Two other studies in this group specifically highlighted the more lipophilic beta-blockers as having a notable association with depression, whereas four other studies reported a more general class effect of beta-blockers​[7,9,16,33,41]​. Sørensen et al. reported that, compared with the general population, patients taking lipophilic beta-blockers are at 2.7 times increased risk of mortality associated with suicide in the first 12 months of use, although there are numerous confounding factors in the lipophilic beta-blocker cohort, which may explain at least part of this effect​[5]​. Li et al. conducted a systematic review, followed by a network meta-analysis of five studies involving 263,025 patients, and found that when compared to diuretics, beta-blockers, calcium channel blockers and angiotensin antagonists were all associated with an increased risk of depression​[7]​. It is not clear whether hypertension was the indication for all drugs and therefore there is a possibility that the findings are dependent on other underlying confounders that are also predictors of depression.

Among the ten studies that found no substantial evidence of any increased risk of depression associated with beta-blockers was a large systematic review and meta-analysis by Riemer et al., which identified 285 randomised controlled trials of 24 beta-blockers​[14]​. The authors highlighted that, compared with placebo, reports of fatigue/tiredness and unusual dreams were increased in users of beta-blockers and suggested that these specific adverse effects may, in many cases, be misinterpreted as depression. The authors of at least two other studies also make this point and, given the relative subjectivity of depression scoring and diagnosis, even guided by DSM-5 or ICD-10 criteria, there is likely to be considerable variability in the differential diagnoses between physicians and even, for each physician, between patients​[38,39,45,46]​.

Of the 13 studies that considered calcium channel blockers, 3 reported an increased risk of depression​[7,33,41]​, 2 reported a decreased risk​[8,13]​, and the remainder found no substantial evidence of any impact on the risk of depression. Those reporting a decreased risk found a similar effect for most antihypertensives. Kessing et al. report the findings of a Danish population-based study to investigate 41 antihypertensive drugs, which found that none of the drugs increased the risk of depression, while several individual drugs were associated with a decrease in depression diagnoses or antidepressant use for each category of calcium channel blockers, drugs that target the RAS and beta-blockers (including propranolol)​[8]​.

The risk associated with drugs that target the RAS appears to be inconclusive, and the majority of studies did not find any substantial evidence associating the use of these drugs and depression. Three studies reported a decrease in the risk for both ACEIs and ARBs compared with other agents or placebo​[8,13,33]​, whereas two studies found an increased risk for ARBs​[7,41]​ and three studies reported an increased risk associated with ACEIs​[16,41,42]​. The magnitude of these effects was small but statistically significant in all studies reporting an association.

For diuretics, seven studies did not find substantial evidence of any effect, although one study reported an increased risk of depression​[41]​, with one further study reporting a decreased risk​[13]​.

To summarise, when pooling the findings from a range of published studies, none of the drugs used to treat hypertension appear to be entirely absent of risk for either depression or symptoms that are associated with depression. However, there is a greater body of evidence covering beta-blockers and almost half of these studies associate this class of drugs, and particularly the more lipophilic beta-blockers, with increased risk of depression. We must be extremely cautious when interpreting these findings as there are many other important variables that may have influenced the outcomes of these studies and that, in turn, influence risk prediction for depression based on the published data. Our observations of many of these factors are discussed in more detail later in this article, but it is important to first consider the similarities and differences between the antihypertensive drug classes which may affect the respective depression risk, in addition to how these agents are prescribed to manage hypertension.

Angiotensin-converting enzyme inhibitors and angiotensin receptor blockers

For the first step of the treatment of hypertension, NICE recommends the use of ACEIs or ARBs in patients aged under 55 years for some categories of patients​[6]​. There is evidence that chronic activation of the RAS may contribute to neuroinflammation, leading to both neurogenic hypertension and increasing risk of psychological and neurological disorders, including depression (where emotion pathways are involved) and dementia​[25,27]​. The arm of the RAS responsible for the proinflammatory effects is known as the classical RAS, which is balanced by the counter-regulatory RAS, which has a neuroprotective role.

In both arms of the RAS, angiotensinogen is cleaved by renin to produce angiotensin I, which in turn is cleaved by ACE to angiotensin II. Through the type 1 angiotensin II receptor (AT₁R), angiotensin II promotes vasoconstriction and inflammation, as part of the classical RAS pathway. But in the counter-regulatory RAS pathway, angiotensin II acts through another receptor type, known as the type 2 angiotensin II receptor (AT₂R), to promote vasodilation and reduced inflammation. At the same time, angiotensin II is cleaved by ACE2, forming angiotensin-(1-7) peptide, which also promotes vasodilation and reduced inflammation via the Mas receptor​[25]​.

Inhibition of ACE counters the chronic activation of the classical RAS pathway, whereas ARBs target angiotensin II AT₁ receptors, thus inhibiting the classical arm of the RAS. Both mechanisms lead to vasodilation, reduced secretion of vasopressin and reduced production and secretion of aldosterone, combining overall to reduce blood pressure. However, it is not clear if the action of ACE inhibitors and ARBs on the peripheral RAS is replicated in the central RAS to exert any neuroprotective effects, although it is known that several of these antihypertensives are active in the CNS (e.g. perindopril, candesartan)​[47,48]​.

Calcium channel blockers

The NICE guideline for hypertension recommends the use of calcium channel blockers in other categories of patients for step 1 of treatment, rather than using an ACEI​[6]​. There is evidence from genetic data that dysfunctional L-type calcium channels are associated with an increased risk of depression and other psychiatric disorders, and that intracellular calcium dysregulation is apparent in depressed patients​[49]​.

Following activation of the voltage-gated calcium channels in the cell membrane, the entry of calcium into the cell is vital for essential cellular processes, including neurotransmitter release, hormone release and gene expression. A dysfunctional calcium channel will disrupt this sensitive signalling, as it may have a lower or higher activation threshold, leading to consequences for the normal functioning of the cell​[26]​. In CNS pathways responsible for mood and emotion, this may manifest as a set of symptoms characterising one or more psychiatric disorders.

This has implications for the use of the groups of drugs known as dihydropyridine calcium channel blockers in patients at risk of or diagnosed with depression, and the prescribing of CCBs has been associated with a decreased risk of depression in some small studies. For many years, despite the more mainstream, evidence-based approach to managing bipolar disorder with lithium and antiepileptics, calcium channel blockers have sometimes been used where tolerability to other agents has been poor​[50]​. Clinical evidence supporting this approach is scant, but recent genetic studies have identified possible connections between L-type calcium channels and the pathology of bipolar disorder​[51–53]​. The effects of Verapamil and diltiazem on depression have been investigated in several studies and it is possible that all calcium channel blockers (dihydropyridines, phenylalkylamines and benzothiazepines) may be associated with changes in psychiatric function, but further studies are needed to explore this relationship​[54]​.

Diuretics

NICE does not recommend thiazide diuretics as part of step one intervention unless one of the first-line options is not tolerated, but this class of drugs is one of the recommended escalation options at step two. There is very little evidence pointing to thiazide diuretics being implicated in depression and the presumed mechanisms of action of thiazides and thiazide-like agents in reducing blood pressure, or their pharmacodynamic properties, are not easily extrapolated to the possible explanations of the underlying pathophysiology of mood disorders. Low-dose thiazide diuretics inhibit the resorption of Na⁺ and Cl^– ions by blocking the Na⁺/Cl^– transporter in the distal convoluted tubules of the kidney, and this results in a greater loss of water in the urine. The reduced overall volume in the blood vessels reduces outward pressure on the vessel walls, thereby producing an antihypertensive effect​[55]​.

Beta-blockers

NICE does not recommend the use of beta-blockers before step four of a treatment strategy for resistant hypertension, and only then if the patient is at risk of (or has) hyperkalaemia and would not therefore tolerate spironolactone(6). Beta-blockers have fallen out of favour in the management of hypertension because evidence shows they are less effective than drugs that target the RAS, calcium channel blockers and diuretics at preventing cardiovascular events, although they are more effective than placebo​[56,57]​.

Beta-blockers are competitive antagonists of the adrenaline and noradrenaline binding sites of the beta-adrenergic receptors in the sympathetic nervous system. Some beta-blockers are non-selective and will block all three sub-types of beta receptor, whereas other agents will more selectively block activity at β₁, β₂ or β₃ receptors. The target for beta-blockers in hypertension management is β₁ receptors, which are found predominantly in the heart and kidneys​[58]​.

The mechanism by which beta-blockers induce an antihypertensive effect is likely multifactorial. Beta-blockers are known to have negative chrono- and inotropic effects, leading to a reduced cardiac output and therefore possibly reduced pressure on the vessel walls. The action of beta-blockers on the sympathetic nervous system (for those agents that can enter the CNS) and the direct effect of systemic beta-blockers on the kidney leads to a reduction in renin output and a consequent decrease in aldosterone secretion. This causes more Na⁺ and water to be excreted in the urine, decreasing blood volume and pressure​[59]​.

Beta-blockers have been implicated in depression almost since they were first used more than 50 years ago, although there remains no consensus on whether or not this effect is real. Those beta-blockers that are more lipophilic, such as propranolol, have a greater ability to diffuse through the blood–brain barrier and can therefore enter the CNS more readily compared with hydrophilic beta-blockers, such as atenolol​[60]​. Lipophilic agents are therefore more likely to be associated with central effects such as fatigue, sleep/dream disorders or depression​[61]​. Beta-blockers disrupt rapid eye movement sleep and can therefore cause insomnia, which in turn leads to poor concentration and attention, impaired performance and even mood disturbance during the daytime​[62,63]​. This impact on sleep and disruption to the sleep/wake cycle is the mechanism that some have proposed connects beta-blocker use with reports of depression​[32]​.

Other antihypertensives

Clonidine, α-methyldopa and moxonidine are centrally-acting antihypertensive drugs. Clonidine is an α₂-antagonist, suppressing noradrenaline and renin release, and is not routinely used in modern hypertension management because sudden withdrawal can lead to a hypertensive crisis​[64]​. α-methyldopa is also an α₂-antagonist and is used, off-label, for managing hypertension in pregnancy, but can cause fluid retention with long-term use​[65]​. Moxonidine causes its antihypertensive effect through activity at the imidazoline receptor I₁. It is sometimes, but rarely, used in resistant hypertension or where other antihypertensives are not suitable​[66]​.

Guanethidine is obsolete in modern hypertension management, except in a hypertensive emergency — although other agents are usually preferred. It works by preventing the release of noradrenaline from nerve endings in the sympathetic nervous system and can lead to a rapid antihypertensive effect, but postural hypotension is common​[67]​.

Hydralazine, sodium nitroprusside, minoxidil and nitrates are all vasodilators that have strong and rapid antihypertensive effects. Several other drugs, including sildenafil, tadalafil and ambrisentan, are licensed for the treatment of pulmonary arterial hypertension. All of these agents are initiated only in specialist cases​[68]​.

NICE states that α₁-antagonists (e.g. prazosin, doxazosin, alfuzosin) may be introduced at step four in the management of resistant hypertension as an alternative to beta-blockers, or patients with hypertension may be prescribed α₁-antagonists for another indication, such as benign prostatic hyperplasia​[69]​. These agents have vasodilator properties and can induce a rapid antihypertensive effect after the first dose. α₁-antagonists act by blocking the action of noradrenaline at the α₁ adrenergic receptors in the vascular smooth muscle, preventing vasoconstriction.

Consideration of other contributing factors and uncertainties 

It is reported that around 20% of patients with ischaemic heart disease (IHD) have major depression, and many more present with symptoms that are often associated with the condition, such as fatigue or unusual dreams, independent of antihypertensive use​[70]​. These important factors are likely to contribute to the association between antihypertensives and depression, and raise important questions about how depression is diagnosed and underlying causal factors.

The psychological burden of living with long-term illnesses or risk factors can itself be the cause of a patient developing symptoms of depression or anxiety, and many patients will have a history of mood disorders that may re-emerge when triggered by a new or worsening comorbidity, hospitalisation or other factors associated with a change to their quality of life​[71,72]​.

For many years, beta-blockers (especially propranolol) have been used effectively to help patients manage their anxiety by controlling the somatic symptoms associated with sympathetic nervous system activity, such as trembling, tachycardia and sweating. There is a strong, recognised association between anxiety and depression, and these disorders often manifest as comorbidities, which poses a problem when trying to understand the role of beta-blockers in raising the risk of mood disorders when other risk factors are already present​[73]​.

It is also important to acknowledge that depression itself is a predictor of heart disease and stroke, with an increased incidence of hypertension in patients who are depressed​[74–76]​. We are therefore faced with the age-old ‘chicken-and-egg’ problem when we further confound this relationship with treatments to control blood pressure.

Reduced adherence to antihypertensive treatment is recognised to be a predictor of depression and is an important confounding factor that should be controlled wherever possible in studies exploring this relationship​[77]​. Interestingly, depression does not appear to be a predictor of statin intolerance — a treatment for another silent risk factor for IHD​[78]​. What is not clear is whether poor adherence to antihypertensives could be a reliable predictor of depression, although uncontrolled hypertension does appear to have a detrimental effect on depression​[79]​.

There are also limitations to the evidence available to us when trying to understand the relationship between adherence and depression. Although reaching different conclusions, both Li et al. and Riemer et al. reported moderate to considerable heterogeneity between studies, reducing the reliability of their findings​[7,14]​. Earlier in this review, we considered the relatively subjective nature of depression diagnosis and we cannot rule out the likelihood of unintended clinician bias when attempting to differentiate between chronic sleep disorders, for example, and depression. It is also likely that there is some degree of reporting bias and publication bias, which may disproportionately add weight to studies reporting a positive correlation between antihypertensive use and depression​[80]​.

Conclusion and future perspectives

While the relationship between antihypertensive use and depression remains unclear, it is important that clinicians continue to work with patients to manage hypertension, for which there is substantial evidence of increased risk of incident stroke or myocardial infarction. It is prudent to consider the risk of depression for each individual based on their medical history and act accordingly, including the patient in the discussion about their treatment options and taking their personal goals and choices into account. The most effective management of hypertension is likely to be when clinical guidelines are followed, the most appropriate dose is prescribed for the patient, and adverse effects are being monitored.

Further studies that investigate the effects of individual antihypertensives are needed, since the physicochemical and pharmacological properties of agents within each class differ sufficiently that there is varying CNS exposure, a varying degree of specificity of receptor activity and, therefore presumably, varying risk of inducing symptoms associated with depression or other psychiatric conditions. Well-designed prospective trials in primary care could provide useful evidence, and large-scale epidemiological studies with carefully considered co-variables would help to refine understanding of the risk.