Theoretical Implications and Clinical Support for Heart Rate Variability Biofeedback for Substance Use Disorders

Abstract

Heart rate variability (HRV) biofeedback is an accessible, cost-effective intervention that has demonstrated clinical value for numerous physical and mental health conditions; however, research on HRV biofeedback in substance use disorders (SUD) is in its nascence. We argue that HRV biofeedback may be particularly beneficial as an adjunct treatment for SUD by targeting bodily systems that are known to be disrupted by chronic substance use and are not historically the focus of psychosocial or pharmacological SUD treatment approaches. In addition to advocating for HRV biofeedback applications in SUD, we identify several gaps within the existing experimental literature, and propose new studies that could accelerate understanding of how and for whom HRV biofeedback is most likely to promote positive behavior change.

1. Introduction

Prominent behavioral and neurobiological models of addiction [1–3] have informed the development of a range of substance use disorder (SUD) interventions designed to augment self-regulation using “top-down” cognitive and behavioral strategies. These psychosocial treatments have been met with enthusiasm but appear useful only in certain individuals and under certain conditions [4]. This article considers a complementary approach to SUD treatment: Directly intervening in “bottom-up” physiological pathways that also are central to self-regulation. The notion of developing mental health interventions that directly target bodily systems is not new. Exercise-based interventions, for example, reduce a range of psychological symptoms and are promising adjuncts for treatment of SUD [5]. Herein, we consider the process of heart rate variability (HRV), an objective and quantifiable physiological biomarker of heart-brain pathways, as a key participant in physical and psychological health [6**]. We then synthesize the direct and indirect evidence that supports HRV-focused interventions as valuable adjuncts in the treatment of SUD.

2. Heart rate variability and substance use

HRV provides a window into how the continually changing external environment and internal milieu mediate cognitive-emotional experience and physiological responses [7, 8*, 9]. A vast literature supports HRV as a robust biomarker of health, wellness, and stress resilience by showing a less variable heart rate among those less able to adaptively respond to their world. Acute and chronic cigarette smoking, electronic cigarette use, as well as passive smoke exposure diminish HRV, possibly in a dose-response manner [10*–12]. Similar to nicotine, consuming alcohol initiates a quick cascade of cardiovascular adaptations [13], with acute alcohol use reducing resting state HRV and heavy chronic alcohol use causing persistent HRV dampening, a relation that may also be dose-dependent [14]. There is evidence that other psychoactive substances (e.g., marijuana, opioids, and cocaine) reduce HRV acutely [15], but the long-term effects and differences between individuals with versus without a SUD are less studied. One preliminary study reported lower HRV in individuals with a heroin use disorder compared to healthy controls [16], but more research is needed to determine the unique and common effects of drugs of abuse on HRV.

Existing research, however, does suggest that the effects of substance use on HRV may be reversible, at least in part. The adverse effects of cigarette smoking, smoke exposure, and nicotine on HRV appear to rebound during abstinence. HRV appears to increase following the transition from smoking to nicotine replacement therapy patch [17, 18] and upon complete discontinuation of nicotine replacement [19], although full recovery of HRV may not occur [20]. Some studies of individuals with alcohol use disorders suggest that HRV increases after 4–6 months of abstinence, whereas other studies report persistently lower HRV during short and long-term abstinence when compared to individuals without an alcohol use disorder [14]. The evidence in support of HRV recovery when use is reduced or stopped is limited, both in terms of the number of studies and their sample sizes, but is nonetheless striking as these studies did not aim to alter HRV; thus, the observed HRV changes appear to have simply co-occurred with cessation of substance use. Next steps include studies that link such cardiovascular changes to changes in mental and physical health outcomes and identify individual difference factors that predict who is most likely to demonstrate these improvements.

3. Heart rate variability biofeedback as an intervention for substance use disorders

An important feature of HRV as a biomarker is its malleability. HRV biofeedback involves learning to pace breathing at a rate that matches the real-time phasic peaks and troughs observed in continuous displays of one’s respiration and beat-to-beat heart rate data. Section 4 elaborates the underlying physiological mechanisms of respiratory sinus arrhythmia (RSA) and baroreflex stimulation that, when synchronized, produce maximum increases in HRV. There are validated, manualized research protocols for five [21] and ten [22] session HRV biofeedback interventions. There are also smart phone applications (apps) available to perform HRV biofeedback virtually anywhere using the phone’s camera as a photoplethysmograph to detect blood volume changes in the fingertip, and proxies of HRV biofeedback using breathing pacers. Thus, by manipulating breathing rate and rhythmicity, it is possible to directly intervene to increase HRV in clinical and control samples that vary in age, health, and diagnosis [9]. Added value of these interventions comes from their ability to instantaneously lower blood pressure and enhance heart rate responses to changes in blood pressure.

Preliminary studies have examined HRV biofeedback as an adjunct to inpatient treatment-as-usual [TAU]) for SUD. One study randomized 48 adults (ages 25–59; > 5 days of abstinence) receiving inpatient treatment for alcohol use to a 2-week, 6-session HRV biofeedback intervention or treatment as usual (TAU) [23]. HRV biofeedback participants evidenced significantly decreased craving from pre- to post-treatment; both groups showed significantly decreased craving 3- and 6-weeks post-treatment compared to pre-treatment. Anxiety decreased significantly from pre-treatment to 3- and 6-weeks post-treatment, in the HRV biofeedback group only [23]. One year later, a non-significant trend towards greater abstinence in the HRV biofeedback group was found [24]. A separate study recruited 48 young men (ages 20–25; > 72 hours of detoxification from alcohol and other drugs) receiving private inpatient substance use treatment to participate in a 3-week, 3-session HRV biofeedback intervention. A non-significant, small effect size decrease in alcohol and drug craving was observed in the biofeedback versus TAU-only group [25]. Interestingly, higher pre-treatment HRV was associated with greater decreases in craving in the TAU group only. The authors speculated that HRV biofeedback dissociated the relation between HRV at treatment entry and changes in craving such that it mitigated the disadvantage of having low HRV at treatment entry for craving reduction. The link between HRV biofeedback and craving was further explored in a non-randomized waitlist-control trial in a college relapse prevention housing sample (N = 46) using a 12-week, 8-session protocol [26]. Significant reductions in craving were found during the biofeedback only, although the difference in simple slopes between biofeedback and control conditions did not differ in this modestly sized sample. There were substantial within and between person differences in craving levels over time and individuals with elevated craving at study entry appeared to experience the greatest reductions in craving. A multilevel reanalysis of these data [27] found that within-person increases in depression symptoms across weeks significantly attenuated, and age and daily practice of > 12 minutes significantly enhanced, HRV biofeedback effects on craving, accounting for 20.5% of the variance in craving changes.

The use of inpatient and abstinent samples in these studies precluded the opportunity to assess whether biofeedback added to TAU led to concomitant substance use reductions. Perhaps the most compelling observation to date, also the subject of a recent systematic review [28**], is that HRV biofeedback (and paced breathing itself) may be useful for promoting decreases in craving, a cardinal feature of SUD and relapse, although larger and better controlled studies clearly are needed. Methodological differences including population, biofeedback protocol, dose, design, and outcome measures make it challenging to draw conclusions from these studies. More definitive studies of HRV biofeedback in outpatient treatment settings wherein physiological changes can be directly mapped onto substance use behaviors are needed. Nonetheless, the existing studies add to the growing evidence that both physical and mental health problems respond to direct body interventions [5, 29**, 30].

4. Heart rate variability biofeedback mechanism of action

The goal of HRV biofeedback is to slow breathing to a rhythmic pace that aligns two mechanisms that control heart rate: RSA (i.e., increases and decreases in heart rate that occur during inhalation and exhalation, respectively) and the baroreflex (i.e., increases and decreases in heart rate that are linked to countervailing changes in blood pressure) [31, 32]. This creates a resonance effect wherein heart rate oscillations are substantially amplified, baroreflex sensitivity is increased, and clinical benefits are obtained [e.g., 29, 33]. There is evidence that resonance breathing, the purported active element of HRV biofeedback, affects baroreceptor-mediated afferent traffic as well as vagal efferents, blood flow to internal organs, and inflammation [32, 34]. Studies also are underway to characterize the neurological underpinnings of resonance breathing’s clinical value to addiction treatment. A recently completed study suggests that resonance breathing decreases activation of visual cortices and increases activation in brain stem and medial prefrontal areas during cue exposure, consistent with a devaluation of automatic, sensory input and enhancement of higher-order, cognitive processing (Bates et al., unpublished results).

The mechanism by which HRV biofeedback affects behavior is not completely clear, but its direct actions on the viscera and its activation of afferent pathways implicate bottom-up pathways that incorporate visceral reactions into holistic brain responses. Such an effect would offer a novel target in which to intervene in both the initiation and maintenance of substance use. In addition to top-down control, bottom-up influences may affect attention bias and/or amplify visceral reaction to cues [9]. A large literature has illustrated attentional bias to salient substance use cues in individuals with (or at high risk for) SUD [36–38]. Critical examinations, however, suggest such attentional biases are contextually mediated by factors such as in-the-moment craving and motivational states [39]. Visceral arousal is another putative mediating factor. In a study of inpatients with alcohol dependence, attention bias to alcohol cues was associated with HRV reactivity to stress-primed alcohol cues [40] and greater reactivity was linked to greater likelihood of relapse by the 6-month follow-up [41]. Moreover, unplanned or maladaptive substance use, when negative consequences are anticipated or conscious intentions are to not use, can occur when decision-making is guided by bottom-up visceral reactions, rather than top-down executive functioning [35].

In addition to these roles, bottom-up visceral reactions may serve to activate and/or interact with cognitive-affective vulnerabilities, such as anxiety sensitivity and distress intolerance. These vulnerabilities reference one’s perceived sensitivity to, and subjective or invivo tolerance of, cardiovascular changes (e.g., self-report versus behavioral indices of distress tolerance; Leyro et al. [43]). They are purported to contribute to negative reinforcement-motivated substance use in a relatively automatic and inflexible manner, particularly within the context of acute affective and physiological distress [42, 43]. For example, anxiety sensitivity and distress tolerance may contribute to how visceral states provoked by nicotine deprivation or other biological stress paradigms (CO₂ inhalation, speech preparation) affect subjective (withdrawal severity, smoking reward) and objective (smoke puff volume) measures of smoking [44–46**] in daily smokers. These effects may extend to visceral states associated with psychopathology; distress tolerance in the context of hyperarousal symptoms of posttraumatic stress has been found to influence alcohol consumption in a community recruited sample [47]. Scant work has examined physiological stress reactivity in this area; however, one study showed distress tolerance moderated the relation between physiological reactivity to psychosocial stress and alcohol use problems in college women [48*]. Further studies characterizing the intersection of subjective distress and discomfort, physiological reactivity, and behavioral responses (i.e., substance use) to stress are needed to elucidate whether treatment adjuncts targeting bottom-up processes can reduce relations between cognitive-affective vulnerabilities and substance use.

In sum, there are several possible avenues by which therapeutic benefits of HRV biofeedback arise: direct dampening of cardiovascular arousal (e.g., increased HR and blood pressure), indirect influences on decision-making, motivation, craving, and attention allocation, and/or interactions with sensitivity to visceral discomfort and unease. Theoretical and emerging empirical work suggests HRV biofeedback may influence each of these constructs. Robust experimental paradigms to study these constructs at behavioral and neural levels will help elucidate important clinical mechanisms.

5. Conclusions

Addiction remains a highly intractable health problem, although the number and nature of SUD treatment strategies continues to expand. There appears to be a general movement towards multifaceted treatment approaches that embrace empirically-supported psychological approaches (e.g., cognitive behavioral therapy, acceptance and commitment therapy, mindfulness-based relapse prevention, contingency management) in combination with pharmacotherapy and transdiagnostic health promoting activities (e.g., yoga, exercise). HRV biofeedback is a feasible adjunct to traditional and complementary SUD interventions and may synergistically address dysregulated body systems that are a consistent feature of SUD.

HRV biofeedback has several qualities that recommend future research and clinical dissemination. The first is practical. HRV biofeedback is relatively easy to teach, learn, and use. It is low cost and, with technological advancements such as HRV biofeedback apps for smart phones, there are few barriers to wide dissemination. In addition, proxies of HRV biofeedback, such as using an app or clock as a visual or auditory signal to pace the inhalation/exhalation cycle to ~0.1 Hz (6 breaths per minute), can be performed strategically both within and outside the lab or clinic, further reducing implementation barriers. Systematic analysis of various mobile phone apps to implement HRV biofeedback is currently unavailable, however, clinical trials are in progress (ClinicalTrials.gov ID: ). Such studies may provide novel insights into the value of HRV biofeedback as the use of mobile apps allows biofeedback to be performed strategically in everyday situation thereby possibly interrupting cognitive-emotional cascades that often characterizes high risk. The most common negative side effect, when resonance breathing is not properly performed, is mild lightheadedness that dissipates upon return to normal breathing. Moreover, because of its safety and low demand characteristics, HRV biofeedback interventions may prove particularly valuable in elderly, disabled, and physically incapacitated populations.

We suggest that there is much to be gained from better understanding the natural synergies between HRV biofeedback and other SUD interventions. For example, several forms of meditation employ slow diaphragmatic breathing, and breathing frequencies during meditation, even by inexperienced meditators, are often at or near that of resonance breathing [49]. Furthermore, HRV biofeedback may work via some paths that are addressed by pharmacotherapy to support smoking cessation. Daily smokers who were administered guanfacine, an α2a -adrenergic agonist, compared to placebo, showed increased HRV following stressful imagery, and those who showed post-stress increases in HRV also demonstrated less subsequent smoking behavior [50]. Thus, pharmacotherapies that facilitate autonomic recovery to stress and/or protect against autonomic failures may promote positive treatment outcomes. HRV biofeedback has a potential advantage of eliciting such physiological changes directly, compared to medications that come with added costs, physician responsibilities, and undesirable side effects. It is thus surprising that, to our knowledge, HRV biofeedback has not been performed with smokers. We posit that the field of addiction would benefit from studies of shared and distinct mechanisms of action in alternative SUD interventions including the adjunctive role of resonance breathing training, meditation and mindfulness practices, and certain pharmacotherapies, as well as SUD treatments as usual.

Finally, it is worth considering a potential value of HRV biofeedback for prevention and early intervention in substance use, as well. Low HRV predicted escalation in cigarette smoking in a longitudinal study of adolescents [51]. Vagal withdrawal, i.e., a decrease in parasympathetic activity, in response to a laboratory stressor was associated with increased likelihood of (and time to) smoking and smoking reinforcement [52]. Moreover, favorable long-term smoking cessation outcomes following a mindfulness-based intervention were associated with pre- to post meditation increase in HRV [53]. More direct research is clearly needed, but these findings are sufficiently consistent with the abundance of evidence that shows that low HRV is a risk factor for a broad spectrum of mental and physical health problems. Thus, bolstering HRV earlier in life, prior to the onset of SUD or other persistent mental health conditions, has transdiagnostic appeal and would be low-cost and low-risk. In a field where prevention and treatment programs, at best, are effective sometimes, for some individuals, and under some circumstances, such adjunct interventions are worthy of greater examination.

It is clear that much work still needs to be done. Rigorous studies are needed to address questions of dose (i.e., duration of intervention, need for daily practice vs ad libitum), subgroups for whom it may be of most benefit (e.g., individuals with greater attention bias toward substance cues or those with distress intolerance), added value to other treatment approaches, and potential role in preventive interventions. It is also important to reiterate that the majority of research examining the relation between substance use and HRV has focused on nicotine and alcohol. Additional research is needed to examine HRV in persons who use other classes of substances (e.g., opiates, cocaine, cannabis), given heterogeneity in pharmacological mechanisms, patterns of use, and consequences, as well as high rates of comorbidity (i.e., polysubstance use). Nonetheless, evidence to date supports the value of HRV biofeedback as complementary to existing SUD treatment strategies by dampening bottom-up visceral reactivity concurrently with psychosocial treatments aimed at enhancing conscious-cognitive control.

Highlights:

• Substance use disorders affect bodily systems central to regulatory processes

• Heart rate variability is associated with substance use onset and maintenance

• Heart rate variability biofeedback may directly improve self-regulation

• Heart rate variability biofeedback warrants consideration as a treatment adjunct for substance use