Anxiety and cardiovascular risk: Review of Epidemiological and Clinical Evidence

Abstract

An increasing body of evidence suggests that anxiety is an independent predictor of adverse cardiovascular (CV) events. Individuals with high levels of anxiety are at increased risk of coronary heart disease, congestive heart failure, stroke, fatal ventricular arrhythmias, and sudden cardiac death. Anxiety following a major cardiac event can impede recovery, and is associated with a higher morbidity and mortality. This linkage between anxiety and CV disease is further corroborated by evidence suggesting that treatment of anxiety may improve cardiac symptoms. However, the mechanisms underlying the observed associations are not entirely delineated. Several intermediary mechanisms have been suggested, including sympathetic activation, impaired vagal control, reduced heart rate variability, stimulation of the hypothalamic–pituitary axis, hyperventilation-induced coronary spasm, oxidative stress, increased inflammatory mediators, and unhealthy lifestyle. There is a need for several clinical trials to explicate the complex associations between anxiety and CV disease, which may be compounded by the involvement of other psychosocial factors. In this review, we examine the epidemiological evidence for the association between anxiety and CV disease, and discuss the proposed mechanisms that may be responsible for this association.

BACKGROUND

Coronary heart disease (CHD) is the leading cause of death and disability in the United States (US) and other developed countries [1]. By the year 2020, CHD is projected to be the leading cause of death, not only in the developed world but also in the developing world [2–4]. It is well known in the psychological literature that negative emotions such as anxiety are common among patients with cardiovascular (CV) symptoms, but few studies have examined associations between these two conditions.

Anxiety is defined as an unpleasant emotion that is triggered by anticipation of future events and memories of past events, and it could manifest in different forms (panic disorder, phobic anxiety, generalized anxiety, anxiety reactions, and chronic anxiety) [5, 6]. Anxiety affects about 24.9% of the population at some time in their lives [7] with reported worldwide prevalence of 16.6% [8]. Studies suggest that 38–70% of all persons with congestive heart failure (CHF) have some form of anxiety, and older adults with CHF report anxiety levels that are estimated to be 60% higher than those without such symptoms [9–13]. Among independently living women, our group observed a greater likelihood for highly anxious women to report CV symptoms [14].

Although the association of traditional CV risk factors with adverse outcomes has been well delineated [15], recent research indicates that the impact of psychosocial distress on CV morbidity and mortality is almost equal to the impact of demographic characteristics (age, sex, and race) and risk markers (smoking, alcohol, obesity, diabetes, dyslipidemia, and hypertension) [16]. Despite this observation, only a few studies have focused on the impact of anxiety and other psychosocial factors on CV outcomes. Underappreciation of the role of psychosocial factors in the development and progression of cardiac diseases may be one reason why they remain the number one cause of death in most developed countries. Although most of the studies exploring the relationship between CV disease and negative emotions have focused on depression, most patients with confirmed or suspected CV disease have some degree of anxiety [9–13]. Anxiety following a major cardiac event can impede recovery, and is associated with higher morbidity and mortality [17–21]. Although several studies have found that anxiety is related to increased mortality [17, 18, 22–25], some studies have found no difference or even a decrease in morbidity and mortality following myocardial infarction (MI) [26–29].

The evidence suggests that treatment of psychosocial distress may decrease CV events [30–33]. However, the mechanisms underlying the observed associations are not entirely delineated. Several intermediary mechanisms have been suggested, including sustained sympathetic activation, hyperventilation-induced coronary spasm, oxidative stress, increased inflammatory mediators, and unhealthy lifestyle changes [34–39]. There is a need for several randomized clinical trials to explicate the complex associations between anxiety and CV disease, which may be compounded by the involvement of other psychosocial factors. In this review, we examine the evidence linking anxiety to CV disease, and discuss the proposed mechanisms that may be responsible for this association.

LINK BETWEEN ANXIETY AND CARDIOVASCULAR SYMPTOMS

Evidence from CHD studies

Epidemiological evidence indicates that pathophysiological correlates of negative emotions may contribute to atherosclerosis [3]. However, several studies evaluating the link between negative emotions and cardiac disease have focused on depression [4]. Recent studies have demonstrated a relationship between hostility or anger and measurements of subclinical atherosclerosis [40–43], and have also linked hostility to progression of atherosclerosis during serial coronary angiography [44]. In a population-based sample of 726 men and women who were healthy at baseline, Paterniti and colleagues [45] showed that high levels of sustained anxiety were independently associated with increased progression of atherosclerosis over a 4-year period, as measured by changes in common carotid artery intima media thickness. Among hypertensive patients, White and Baker [46] found a mean increase of 27 mmHg in systolic blood pressure (SBP) and 5 mmHg in diastolic blood pressure (DBP), and an increase of 14 beats/min in the heart rate during the hour of a panic attack. In other studies, anxiety level was positively correlated with SBP [47–49], but not with DBP [50].

Three large-scale, community-based studies including one involving 34000 men showed significant relationships between anxiety disorders and CV death [17, 18, 25]. Haines and colleagues [25] followed 1457 initially healthy men for about 10 years in the Northwick Park heart study. Those with the highest levels of phobic anxiety had a higher risk of fatal CHD than men reporting no anxiety, after controlling for a range of known coronary risk factors [25]. Furthermore, a recent study from a US population showed that anxiety was associated with 60% excess risk of CHD among men and women, an effect that was independent of traditional CHD risk factors [51]. Chronic anxiety appears to increase the risk of incident CHD, with risk estimates from 1.5 to 7, depending on the type of anxiety measure used and the form of the analysis [17, 18, 25, 51–58]. Among healthy individuals, higher levels of anger symptoms were significantly associated with a 1.5- to threefold excess risk of incident CHD over a 5- to 15-year follow-up period [59–63].

In two studies involving patients participating in cardiac rehabilitation programs, anxiety symptoms predicted worse long-term prognosis [19, 20]. As these studies included psychological treatment components during follow-up, it is difficult to interpret the results. In another study, 222 patients who received usual care for 1 year were examined during hospitalization for MI. The results of that study indicated an increase in CV risk associated with anxiety symptoms that was independent of history of major depression and previous MI [21]. Other studies have also shown associations of anxiety with cardiovascular events [23, 64–69], but differences in the measures used, timing of measurement, and variations in sample sizes make it difficult to establish a head-to-head comparison of those findings.

Findings of positive relationships between anxiety and CV disease are not universal. One large study involving over 2000 patients with CAD assessed before routine stress tests found that anxiety, measured with the anxiety subscale of the Hospital Anxiety and Depression Scale [70], was associated with lower mortality rates [27]. Investigators in that study speculated that anxiety might be related to increased tendencies to seek medical attention or alter risk factors [27]. Another study that evaluated 344 patients after MI reported no relationship between either anxiety or depression and 1-year mortality [26]. However, the mortality rate was low (4%) in that study.

Evidence from CHF studies

Evidence supporting a prognostic impact of anxiety on CHF is scant and not entirely consistent. Riedinger and colleagues [71] demonstrated that anxiety was an independent predictor of adverse cardiac events among patients with recent MI and low ejection fraction. In another study, anxiety was directly related to brain natriuretic peptide level [72], suggesting a relationship between anxiety and CHF [72]. However, these findings are not consistent with other studies showing that poor outcomes and rehospitalization were associated with depression, but not with anxiety [73–78]. Friedman et al [73] found an association between anxiety and outcomes among patients with heart failure, but this association was not maintained after controlling for potential confounders. Jiang et al [74] did not find a significant association between anxiety symptoms and mortality in hospitalized patients with heart failure. The reasons for these discrepancies are not completely understood, but they may be related to differences in study population and methodology.

Evidence from psychosocial intervention studies

The impact of psychosocial interventions on CV outcomes represents a new area of interest. Epidemiological evidence suggesting that treatment of anxiety and other psychosocial factors may improve CV symptoms and decrease adverse outcomes corroborate the link between psychosocial risks and CV disease. In the initial successful intervention trial in this arena, the Ischemic Heart Disease study, a unique home-based stress reduction program, showed that treatment was associated with reduction in cardiac events [30]. The Recurrent Coronary Prevention Project Study succeeded in decreasing both type A behavior and negative affect, and also reduced the rates of CV mortality and non-fatal MI by using behavior modification [31]. Furthermore, among patients referred to cardiac rehabilitation programs, Dusseldorp et al [32] observed differential reduction in the odds ratio for mortality and recurrent MI following reduction in psychological stress. In a meta-analysis of 23 randomized controlled trials that evaluated the impact of adding psychosocial interventions to standard cardiac rehabilitation regimens, Linden et al [33] observed that, during the first 2 years of follow-up, lack of psychosocial intervention was associated with greater rates of mortality and recurrent MI.

PROPOSED MECHANISMS EXPLAINING THE LINK BETWEEN ANXIETY CV DISEASE

Mechanisms linking anxiety to CV disease can be categorized into two pathways: neurohormonal and behavioral pathways. Below, we discuss the significance and evidence in support of both pathways.

Neurohormonal pathway

Activation of the sympathetic nervous system, as manifested by impaired vagal control, reduction in heart rate variability [34], elevated levels of proinflammatory cytokines [5], and hypercortisolemia from activation of the hypothalamic–pituitary axis (HPA) [35], is increasingly recognized as a favorable link between anxiety and CV disease [34–36]. Epidemiological evidence indicates that negative emotions have cumulative pathophysiological effects that can ultimately lead to CHD events through damage from persistent activation of the neurohormonal systems and other mechanisms [37–39]. Increased output from the sympathetic nervous system and HPA can induce a wide variety of physiologic responses [22]. Studies have shown that the number of lymphocyte b-adrenergic receptors [79–81] and b-adrenergic-mediated cyclic AMP generation are decreased among patients with anxiety disorder, especially in the context of panic disorder [82, 83]. Aronson et al [80] demonstrated a negative correlation between trait anxiety and the number of lymphocyte b-adrenergic receptors. Among patients experiencing panic attacks, baseline catecholamine and 3-methoxy-4-hydroxyphenylglycol levels appear to be either normal or mildly elevated [84–86]. Following infusion of isoproterenol, a smaller change in the heart rate was observed among patients with panic disorder [87], suggesting that chronic anxiety, by increasing sympathetic outflow, can lead to b-adrenergic receptor downregulation.

Data from the Normative Aging Study, which previously identified an excess risk of CHD associated with anxiety, also showed a link between high levels of anxiety and reduced heart rate variability [34]. With increased anxiety symptoms, there was a corresponding decrease in heart rate variability. This finding supports studies suggesting that the association of anxiety with sudden cardiac death, but not MI [17, 18, 25], may be related to ventricular arrhythmias. Autonomic alteration could involve either increased sympathetic stimulation, which has been linked to the occurrence of arrhythmias and sudden death [88], or impaired vagal control, which has also been linked to increased cardiac mortality [89, 90]. Specifically, reduced vagal control has been linked to impaired vagally mediated baroreflex control of the heart [91]. Such impairment appears to be a particularly important risk factor for sudden death [92, 93]. Watkins et al [94] reported reduced baroreflex cardiac control among patients with anxiety, but prospective work is needed to determine whether this is a common operative mechanism for sudden deaths among patients with anxiety syndromes. Consistent with these findings, we previously observed that insomnia, which has been shown to be a form of chronic stress and sympathetic disorder, is in part responsible for the association between anxiety and CV symptoms [14]. Furthermore, an atherogenic pathway, promoted by recurring activation of the neurohormonal system with subsequent endothelial injury and atherosclerosis, has also been suggested [45].

Behavioral pathway

Various lifestyle behaviors, including unhealthy diet, physical inactivity, and smoking, promote the development and clinical manifestations of CHD [22]. Studies suggest that individuals with anxiety disorders are prone to unhealthy lifestyle behaviors [16, 95]. Negative emotions may exacerbate disease progression or reduce survival, either via direct physiological effects or through reduced compliance with recommended medical regimens [96, 97]. Among patients with CHF, negative emotions are associated with unhealthy lifestyle [98], and have been shown to be predictors of dietary and exercise adherence [11, 99, 100].

Riegel et al [101] suggest that the impact of anxiety on cardiac diseases may be related to the associated poor uptake of self-care behaviors. Anxiety may impair energy, cognition, and motivation to engage in self-care behaviors [96]. Patients with high levels of anxiety have difficulty making lifestyle changes, coping with challenges, and encounter more problems during cardiac rehabilitation [102–104]. Evidence suggests that anxiety is a predictor of poor adherence to lifestyle change recommendations and adherence to medical therapy [105]. Among highly anxious patients with recent hospitalization for acute exacerbation of CHF, adherence to a number of self-care behaviors was extremely poor [11]. In that study, only 9% of patients reported self-monitoring for symptoms of worsening CHF, 14% of patients were weighing themselves daily, 31% could not name even one symptom of CHF, and only 34% of patients were taking all the medications as prescribed. These findings suggest that unhealthy lifestyle behaviors may contribute to cardiac risks associated with anxiety.

A general effect of distress

The clustering of psychosocial factors and the similarity of findings between the three negative emotions and CV outcome raise the question as to whether the effects of anxiety, depression, or anger are unique, or just a general effect of distress. It has been suggested that general distress shared by these negative emotional states is a key factor [106], but little work has been done to tease apart these relationships. Recently, the uniqueness of these three negative emotions as risk factors for incident CHD was explored in a prospective study of a disease-free population, which was followed for 11 years [96]. General stress, anxiety, and depression measured at study entry were each strong predictors of incident CHD over the follow-up period. At the end of the follow-up period, anxiety and depression, but not anger, remained independent predictors. These findings support the need to consider these specific emotions separately. It may also be that general distress in and of itself is an important CV risk factor.

ANXIETY AND TRADITIONAL CORONARY RISK FACTORS

A growing body of evidence indicates that psychosocial factors may be considered as major CHD risk factors in a similar way to the traditional risk factors (hypertension, diabetes, hyperlipidemia, smoking, obesity, physical inactivity, and age). This notion is supported by the multinational INTERHEART case–control study of 29972 patients from 52 countries, which uniquely assessed the impact of eight coronary risk factors and a composite index of psychosocial factors on incident acute MI [16]. Psychosocial distress conferred a greater adjusted relative risk of acute MI than hypertension, abdominal obesity, diabetes, and several other traditional risk factors. In addition, psychosocial distress was a predictor of acute MI independent of ethnic or geographic context [107]. This important finding, along with reports from the Framingham study [95, 108], suggests that psychosocial factors may be equivalent to traditional risk factors in predicting CV events, although more studies are needed in this area.

CONCLUSION

With the limited evidence available, anxiety appears to predict more CV symptoms, poorer functional status, and increased CV events. More experimental and randomized clinical trials are needed to fully understand the relationship between anxiety and CV disease, as well as to determine whether the proposed pathophysiological mechanisms are truly causal. Clinical evaluation of cardiac patients should include assessment of psychosocial factors, most especially anxiety and depression. Further research in this area should concentrate on teasing out the mechanisms by which individual psychosocial factors increase CV risk, and also focus on interventions that decrease both anxiety and adverse CV events.