Blood pressure and muscle sympathetic nerve activity are associated with trait anxiety in humans

Abstract

Chronic anxiety is prevalent and associated with an increased risk of cardiovascular disease. Prior studies that have reported a relationship between muscle sympathetic nerve activity (MSNA) and anxiety have focused on participants with anxiety disorders and/or metabolic syndrome. The present study leverages a large cohort of healthy adults devoid of cardiometabolic disorders to examine the hypothesis that trait anxiety severity is positively associated with resting MSNA and blood pressure. Resting blood pressure (BP) (sphygmomanometer and finger plethysmography), MSNA (microneurography), and heart rate (HR; electrocardiogram) were collected in 88 healthy participants (52 males, 36 females, 25 ± 1 yr, 25 ± 1 kg/m²). Multiple linear regression was performed to assess the independent relationship between trait anxiety, MSNA, resting BP, and HR while controlling for age and sex. Trait anxiety was significantly correlated with systolic arterial pressure (SAP; r = 0.251, P = 0.018), diastolic arterial pressure (DAP; r = 0.291, P = 0.006), mean arterial pressure (MAP; r = 0.328, P = 0.002), MSNA burst frequency (BF; r = 0.237, P = 0.026), and MSNA burst incidence (BI; r = 0.225, P = 0.035). When controlling for the effects of age and sex, trait anxiety was independently associated with SAP (β = 0.206, P = 0.028), DAP (β = 0.317, P = 0.002), MAP (β = 0.325, P = 0.001), MSNA BF (β = 0.227, P = 0.030), and MSNA BI (β = 0.214, P = 0.038). Trait anxiety is associated with increased blood pressure and MSNA, demonstrating an important relationship between anxiety and autonomic blood pressure regulation.

NEW & NOTEWORTHY Anxiety is associated with development of cardiovascular disease. Although the sympathetic nervous system is a likely mediator of this relationship, populations with chronic anxiety have shown little, if any, alteration in resting levels of directly recorded muscle sympathetic nerve activity (MSNA). The present study is the first to reveal an independent relationship between trait anxiety, resting blood pressure, and MSNA in a large cohort of healthy males and females devoid of cardiometabolic comorbidities.

Listen to this article’s corresponding podcast at https://ajpheart.podbean.com/e/anxiety-and-muscle-sympathetic-nerve-activity/.

INTRODUCTION

Anxiety is common within the general population, with an estimated lifetime prevalence for an anxiety disorder of ∼30% (1). Anxiety is associated with several negative health outcomes, including the development of cardiovascular disease (2–5). Two recent meta-analyses reported significant associations between anxiety disorders and hypertension and highlighted a causal role for anxiety in the development of hypertension by integrating prospective studies within the analyses (2, 3). Anxiety can also exacerbate adverse cardiovascular conditions (6), supporting a critical role for anxiety in cardiovascular disease pathology and management.

A suspected mediator of the relationship between anxiety and hypertension (2, 3) is hyperactivity of the sympathetic nervous system. Measurable increases in sympathetic outflow via direct recording of muscle sympathetic nerve activity (MSNA) have been observed in the pathogenesis of numerous cardiovascular disorders, including hypertension (7). Although MSNA has been recorded in anxiety disorders such as generalized anxiety disorder (GAD) (8) and panic disorder (PD) (9–12), these studies surprisingly indicate little, if any, alterations in multiunit MSNA at rest compared with healthy controls, and instead primarily highlight altered sympathetic neural responses to stress (13). Although exaggerated pressor responses to laboratory stress are predictive of future cardiovascular dysfunction (14), the impact of exaggerated MSNA reactivity on future cardiovascular health has not yet been characterized. Conversely, longitudinal changes in resting MSNA are associated with changes in blood pressure (15), and elevated resting MSNA is observed in numerous cardiovascular disease states (7, 16, 17), supporting the utility of understanding differences in resting sympathetic outflow when investigating propensity for future cardiovascular risk in individuals with chronic anxiety.

Despite the lack of evidence supporting altered levels of resting MSNA in diagnosed anxiety disorders, levels of trait anxiety, which assess an individual’s propensity for anxiousness (18, 19), have been shown to correspond to heightened resting multiunit MSNA in individuals with metabolic syndrome (20). However, these findings are complicated by the presence of comorbid metabolic and cardiovascular dysfunction (20). To our knowledge, a relationship between trait anxiety, blood pressure, and resting MSNA has not been established in adults free of overt cardiometabolic disease.

The present study leverages the largest study population to date with concurrent trait anxiety measures, resting blood pressure, and direct recordings of sympathetic neural activity (i.e., MSNA). Importantly, the utilization of trait anxiety assessment helps to circumvent the limitations of viewing anxiety disorders as dichotomous variables, rather than a spectrum of symptom severity, when assessing associations with cardiovascular function. We hypothesized that heightened trait anxiety would be associated with increased resting MSNA and blood pressure in the current sample of healthy adults.

METHODS

Participants

Subjects were analyzed from retrospective (21–23) and ongoing studies at Montana State University, Purdue University Northwest, University of Minnesota, and Michigan Technological University. All studies were approved by the corresponding institutional review boards, and all participants provided written, informed consent to participate in each respective study.

Participants were healthy adults between the ages of 18 and 47 yr. Participants were required to have a body mass index (BMI) < 35 kg/m² and all self-reported as nonsmokers. All participants self-reported the absence of diabetes, as well as any known cardiovascular or autonomic disorders. Participants with stage II hypertension or greater [i.e., systolic arterial pressure (SAP) > 140 mmHg or diastolic arterial pressure (DAP) > 90 mmHg] were excluded. Participants reported no usage of cardiovascular or autonomic medications. All females studied were premenopausal and reported regular menstrual cycles (i.e., 26–30 days in length). Females were tested in the early follicular phase of the menstrual cycle, or low hormone phase if using hormonal contraceptives. On the day of autonomic testing, participants were instructed to fast for at least 3 h before testing, and refrain from caffeine, alcohol, and exercise for at least 12 h. One participant’s blood pressure was a significant outlier (96/44 mmHg) with a standardized residual that exceeded three standard deviations, thus was removed from analysis. A total of 88 participants (52 males, 36 females, 25 ± 1 yr, 25 ± 1 kg/m²) were included in the final analyses.

Study Design

Participants arrived at the laboratories of Montana State University, Purdue University Northwest, University of Minnesota, or Michigan Technological University to undergo autonomic testing. All participants completed the trait anxiety portion of the Spielberger State/Trait Anxiety Inventory (STAI) (19) before the autonomic function test at either the initial orientation session or on the morning of the autonomic testing session. Upon arrival to the respective laboratory on the scheduled day of testing, participants were instructed to sit quietly for at least 5 min, after which three seated brachial blood pressure measurements were taken, separated by 1 min each (Omron HEM-907XL, Kyoto, Japan). In six participants, supine brachial blood pressure measurements were taken instead of seated and used for resting blood pressure quantification in these individuals. Findings were not different with or without these six participants and given that MSNA was the primary outcome variable we included these individuals in the final blood pressure analyses. Immediately following resting blood pressure measurements, participants were brought into the autonomic testing laboratory and positioned supine or semirecumbent on a cushioned tilt table. Measurement of participants’ heart rate (HR, electrocardiogram), beat-to-beat blood pressure via finger plethysmography (NOVA, Finapres Medical Systems, Amsterdam, The Netherlands and Human NIBP Controller, AD Instruments, Colorado Springs, CO), and muscle sympathetic nerve activity (MSNA; 662 C-3 Nerve Traffic Analysis System, University of Iowa Bioengineering and NeuroAmp EX, ADInstruments) were obtained. Participants were then asked to sit quietly for 10 min of nonrecorded rest before data collection to ensure the accuracy of all measured cardiovascular and autonomic parameters. Three additional brachial blood pressure measures were next taken, and the average was used to calibrate the finger plethysmography before recording. Finally, 10 min of quiet, resting autonomic and cardiovascular measurements were collected. In three participants, we obtained 5 min of quality MSNA and heart rate.

Measurements

Spielberger State/Trait Anxiety Inventory.

The STAI (19) is a 40-question assessment used to evaluate both state and trait anxiety, with each section consisting of 20 questions. For the current study, only the trait anxiety portion of the assessment was used to quantify chronic anxiety symptoms and overall propensity for anxiousness. The trait anxiety scale serves as a valid metric of chronic anxiety severity with adequate psychometric properties (19, 24, 25), and has shown intraindividual reliability over time (19). We report on the trait anxiety percentile scores provided by Spielberger et al. (19), which provides a normative percentile rank based on participant age and biological sex. The results of the current study were unchanged when using the raw STAI scores rather than the percentile rankings.

Blood pressure and heart rate.

The average of the three resting brachial blood pressure measurements taken before the autonomic study were used to quantify average systolic (SAP), diastolic (DAP), and mean arterial pressure (MAP) in each participant. Beat-to-beat blood pressure was only used to assess sympathetic baroreflex sensitivity. Average HR from the 5–10 min autonomic session was used to quantify average resting HR.

Muscle sympathetic nerve activity.

Multifiber MSNA was obtained and recorded using microneurography (662 C-3 Nerve Traffic Analysis System, University of Iowa Bioengineering and NeuroAmp Ex, ADInstruments) as previously described by our laboratory (21, 23) and a recent guidelines paper (26). Briefly, the location of the common peroneal nerve was mapped using external electrical stimulation at the level of the popliteal fossa and fibular head to evoke involuntary muscle activation resulting in lateral dorsiflexion. Next, a tungsten recording electrode was inserted percutaneously into the peroneal nerve of the right leg at the level of the popliteal fossa or head of the fibula. A reference electrode was inserted 2–3 cm away from the recording electrode. The electrodes were then connected to a differential preamplifier and then to an amplifier (total gain, 80,000) where the nerve signal was band-pass filtered (700–2,000 Hz) and integrated (time constant, 0.1). The recording of MSNA was confirmed by assessing the nerve response to end-expiratory apnea, as well as the absence of a response to startle (i.e., shout/clap). Once a recording site was obtained, it was not adjusted throughout the duration of the recording.

Data Analysis

All data were imported and analyzed in either WinCPRS (Absolute Aliens, Turku, Finland) or Ensemble (Elucimed, Ltd., Wellington, New Zealand). Regardless of analysis software, R waves of the ECG were automatically detected, manually confirmed, and used to quantify heart rate. Bursts of MSNA were detected automatically based on a signal-to-noise ratio of 3:1 and a burst peak latency of ∼1.3 s from the previous R wave. All bursts marked by the software were manually reviewed and edited by a trained investigator. MSNA was expressed as both burst frequency (BF; bursts/min) and burst incidence [BI; bursts/100 heartbeats (hb)].

The slope of the weighted linear regression of DAP and MSNA during the autonomic baseline was used to quantify spontaneous sympathetic baroreflex sensitivity (sBRS) as previously described (21, 23). DAP values on a beat-to-beat basis were binned into 3-mmHg intervals and were subsequently plotted against MSNA burst incidence. All slopes were r > 0.7 based on recent recommendations (27).

Statistical Methods

All data were analyzed using SPSS statistics version 28.0 (IBM Corp., Armonk, NY). Descriptive statistics were performed to characterize participant characteristics (i.e., anthropometric, trait anxiety, cardiovascular, and autonomic variables).

Multiple linear regression was used to examine the independent relationship between trait anxiety and outcome variables of interest. The primary regression models all included age, sex, and trait anxiety as predictor variables, and either average SAP, DAP, MAP, resting HR, MSNA BF, or MSNA BI as dependent variables. BMI was not included as a predictor variable as it was not correlated with the main outcome variables (MSNA BF, R = 0.069, P = 0.525; MSNA BI, R = 0.089, P = 0.410) and did not contribute significantly when included in the regression models. Multicollinearity between predictor variables was assessed, and none of the predictor variables exhibited significant multicollinearity (i.e., all r < 0.5). All regression models were explored for outliers with standardized residuals exceeding three standard deviations. Finally, all residuals were determined to be normally distributed.

Since a portion of the presented findings were taken from previous studies (21–23), an a priori power analysis was not conducted. However, a post hoc sensitivity analysis was performed to find the minimum effect size that the data set could accurately detect. Multiple regression using three predictors (i.e., age, sex, and trait anxiety), with a total sample size of 88, and a power (1 – β) of 0.8 at an α = 0.05 resulted in a minimum detectable effect size of R² = 0.115. Although sex differences were initially explored, post hoc sensitivity analysis indicated insufficient power to probe for these differences with adequate confidence. For this reason, sex differences were not included in the current data set, but this remains an interest for future work. All data are presented as means ± SE. A significance level of α < 0.05 was used for all statistical analyses.

RESULTS

Subject Characteristics

Demographics, seated cardiovascular, and resting autonomic measures are presented in Tables 1 and 2. sBRS was assessed in 58 (31 males, 27 females) participants because of DAP-MSNA correlations that were positive or below the a priori cutoff (i.e., r < 0.7). sBRS was found to be unrelated to trait anxiety (r = −0.028, P = 0.835) and was not explored in the regression analyses.

Table 1.

Participant characteristics

Variable
n (male/female)	88 (52/36)
Age, yr	25 ± 1 [18–47]
BMI, kg/m2	25 ± 1 [18–34]
STAI-T (raw)	36 ± 1 [20–67]
STAI-T, %	51 ± 3 [3–100]

Values are means ± SE [range]. BMI, body mass index; STAI-T, trait anxiety form of the Spielberger State-Trait Anxiety Inventory expressed as raw values and percentile.

Table 2.

Resting cardiovascular measures

Variable
SAP, mmHg	115 ± 1
DAP, mmHg	70 ± 1
MAP, mmHg	85 ± 1
HR, beats/min	63 ± 1
MSNA BF, bursts/min	20 ± 1
MSNA BI, bursts/100 hb	31 ± 1
sBRS, bursts/100 hb/mmHg	−2.1 ± 0.2

Values are means ± SE; N = 88 (52 males, 36 females) participants for all measures except sympathetic baroreflex sensitivity (sBRS). sBRS was measured in 58 participants (31 males, 27 females). BF, burst frequency; BI, burst incidence; DAP, diastolic arterial pressure; hb, heartbeats; HR, heart rate; MAP, mean arterial pressure; MSNA, muscle sympathetic nerve activity; SAP, systolic arterial pressure.

Correlational Analyses

Table 3 includes correlation coefficients between independent (age, sex, and trait anxiety) and dependent variables (SAP, DAP, MAP, HR, MSNA BF, and MSNA BI). Trait anxiety was significantly correlated with SAP (r = 0.251, P = 0.018), DAP (r = 0.291, P = 0.006), MAP (r = 0.328, P = 0.002), MSNA BF (r = 0.237, P = 0.026), and MSNA BI (r = 0.225, P = 0.035).

Table 3.

Correlation matrix

	Age	Sex	STAI, au	STAI, %	SAP	DAP	MAP	HR	MSNA BF	MSNA BI
Age
R	1.00
P
Sex
R	0.167	1.00
P	0.120
STAI, au
R	−0.086	−0.108	1.00
P	0.424	0.316
STAI, %
R	−0.059	−0.111	0.957	1.00
P	0.583	0.305	<0.001
SAP
R	0.057	−0.480	0.269	0.251
P	0.600	<0.001	0.011	0.018	1.00
DAP
R	0.316	0.080	0.268	0.291	0.377
P	0.003	0.458	0.012	0.006	<0.001	1.00
MAP
R	0.262	−0.175	0.321	0.328	0.754	0.892	1.00
P	0.014	0.104	0.002	0.002	<0.001	<0.001
HR
R	−0.104	−0.020	0.101	0.127	0.140	0.305	0.283	1.00
P	0.336	0.853	0.350	0.239	0.194	0.004	0.007
MSNA BF
R	0.162	−0.195	0.224	0.237	0.293	0.279	0.343	0.410	1.00
P	0.133	0.069	0.036	0.026	0.006	0.008	0.001	<0.001
MSNA BI
R	0.191	−0.208	0.210	0.225	0.273	0.216	0.289	0.138	0.949	1.00
P	0.074	0.052	0.050	0.035	0.010	0.043	0.006	0.199	<0.001

Values are Pearson’s correlation coefficients (R) and two-tailed significance levels (P). DAP, diastolic arterial pressure (mmHg); HR, heart rate (beat/min); MAP, mean arterial pressure (mmHg); MSNA, muscle sympathetic nerve activity quantified as either burst frequency (BF, bursts/min) or burst incidence (BI, bursts/100 heartbeats); SAP, systolic blood pressure (mmHg); STAI, trait form of the Spielberger State Trait Anxiety Inventory denoted as raw arbitrary units (au) and percentiles (%). N = 88 (52 males, 36 females) participants.

Trait Anxiety, Blood Pressure, and MSNA

Multiple regression was performed to predict SAP, DAP, MAP, HR, MSNA BF, and MSNA BI with age, sex, and trait anxiety as predictors. These variables significantly predicted SAP [F(3,84) = 11.521, R² = 0.292, P < 0.001], DAP [F(3,84) = 6.996, R² = 0.200, P < 0.001], MAP [F(3,84) = 8.005, R² = 0.222, P < 0.001], MSNA BF [F(3,84) = 4.085, R² = 0.127, P = 0.009], and MSNA BI [F(3,84) = 4.592, R² = 0.141, P = 0.005], but not HR [F(3,84) = 0.731, R² = 0.025, P = 0.536].

Table 4 demonstrates the independent associations between each of the predictor variables and the outcome variables of interest. Notably, when controlling for the effects of age and sex, trait anxiety was independently associated with SAP (β = 0.206, P = 0.028, Fig. 1A), DAP (β = 0.317, P = 0.002, Fig. 1B), and MAP (β = 0.325, P = 0.001, Fig. 1C). Trait anxiety was additionally independently associated with MSNA BF (β = 0.227, P = 0.030, Fig. 2A) and MSNA BI (β = 0.214, P = 0.038, Fig. 2B).

Table 4.

Regression coefficients

	B	SE	β	t	P
SAP
Age	0.248	0.154	0.149	1.602	0.113
Sex	−10.430	2.026	−0.482	−5.147	<0.001
STAI-T%	0.079	0.036	0.206	2.232	0.028
DAP
Age	0.384	0.117	0.325	3.279	0.002
Sex	0.942	1.537	0.061	0.613	0.542
STAI-T%	0.087	0.027	0.317	3.226	0.002
MAP
Age	0.351	0.109	0.314	3.210	0.002
Sex	−2.791	1.433	−0.191	−1.947	0.055
STAI-T%	0.084	0.025	0.325	3.358	0.001
HR
Age	−0.132	0.147	−0.098	−0.898	0.372
Sex	0.174	1.930	0.010	0.090	0.929
STAI-T%	0.038	0.034	0.122	1.124	0.264
MSNA BF
Age	0.297	0.147	0.209	2.021	0.046
Sex	−3.789	1.926	−0.205	−1.968	0.052
STAI-T%	0.075	0.034	0.227	2.209	0.030
MSNA BI
Age	0.524	0.223	0.242	2.353	0.021
Sex	−6.355	2.923	−0.224	−2.174	0.032
STAI-T%	0.108	0.051	0.214	2.103	0.038

Values are resultant regression values from multiple linear regression analysis; n = 88 (52 males, 36 females) participants in all analyses. B, unstandardized regression coefficient; β, standardized regression coefficient; DAP, diastolic arterial pressure; HR, heart rate; MAP, mean arterial pressure; MSNA, muscle sympathetic nerve activity expressed as bursts/min [burst frequency (BF)] and bursts/100 heartbeats [burst incidence (BI)]; SAP, systolic arterial pressure; STAI-T, trait anxiety form of the Spielberger State-Trait Anxiety Inventory.

Figure 1.

Associations between trait anxiety and blood pressure following multiple linear regression. Trait anxiety was independently associated with systolic (SAP; A), diastolic (DAP; B), and mean (MAP; C) arterial pressure when controlling for the effects of age and sex. Dotted lines indicate the 95% confidence bands of the best fit line. n = 88 (52 males, 36 females).

Figure 2.

Associations between trait anxiety and resting muscle sympathetic nerve activity (MSNA) following multiple linear regression. Trait anxiety was independently associated with resting MSNA burst frequency (BF; A) and burst incidence (BI; B) when controlling for the effects of age and sex. Dotted lines indicate the 95% confidence bands of the best fit line. n = 88 (52 males, 36 females).

DISCUSSION

Chronic anxiety is highly prevalent and contributes to cardiovascular diseases such as hypertension (2, 3) and coronary heart disease (4), although the role of the sympathetic nervous system in this relationship remains equivocal. The present study examined the independent relationship between trait anxiety, blood pressure, and MSNA within a robust population of healthy adults, and we report two key findings. First, trait anxiety was independently associated with resting blood pressure. Second, trait anxiety was independently associated with resting MSNA. To our knowledge, the present study represents the largest sample size to date assessing the link between trait anxiety and resting cardiovascular function in tandem with direct measurement of MSNA via microneurography. These findings offer evidence of a link between heightened trait anxiety, augmented sympathetic outflow, and elevated blood pressure in otherwise healthy adults, highlighting a tangible relationship between chronic anxiety and autonomic function.

Previous studies have investigated the influence of chronic anxiety on blood pressure and MSNA using clinical populations such as GAD (8), PD (9–12), specific phobias (28), as well as trauma-related conditions such as posttraumatic stress disorder (PTSD) (29–31). Despite the prevailing notion that these conditions are associated with global sympathoexcitation, few studies have reported augmented resting multiunit MSNA in these populations. In fact, the only significant elevations in MSNA were reported in individuals with PTSD (31), a condition categorized as a “trauma and stressor-related disorder” rather than an anxiety disorder in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.). Nevertheless, significant overlap between PTSD and anxiety disorders has been reported (32, 33). Furthermore, these elevations in MSNA were present when compared with healthy, noncombat experienced controls (31), but not apparent when compared with combat experienced veterans without PTSD (29, 30).

In contrast, Toschi-Dias et al. (20) explored the impact of clinically diagnosed anxiety/mood disturbance comorbid with metabolic syndrome on MSNA, and reported that individuals who had anxiety/mood disturbance in addition to metabolic syndrome had heightened MSNA at rest compared with healthy controls and individuals with metabolic syndrome alone. Although these findings are informative, the underlying metabolic and cardiovascular characteristics of this cohort make assessment of the independent effects of trait anxiety on MSNA difficult to disentangle. For instance, the individuals diagnosed with comorbid anxiety and mood disturbance had the highest levels of circulating triglycerides (20), a metabolic characteristic that was found to be associated with resting MSNA in a recent meta-analysis, particularly in obese individuals (34).

The present study addresses an important scientific gap by investigating the influence of trait anxiety on neural cardiovascular function in healthy adults while controlling for age and cardiometabolic health through study inclusion/exclusion criteria and statistical analyses. Our findings demonstrate significant and independent associations between trait anxiety, sympathoexcitation, and blood pressure. With 88 healthy participants, this is the largest study to date assessing the relationship between trait anxiety, blood pressure, and directly measured peripheral sympathetic activity (i.e., MSNA). The observed elevations in both blood pressure and MSNA in the absence of sBRS alterations suggest a resetting of baroreflex control of peripheral sympathetic outflow to a higher blood pressure set point in individuals with heightened trait anxiety. Although the central brain regions underlying this resetting are not characterized in their entirety, hypothalamic structures such as the paraventricular nucleus and dorsomedial hypothalamus may play a key role (35–37). Evidence in humans demonstrates a role for the dorsomedial hypothalamus in regulation of MSNA (38). The hypothalamus receives input from structures such as the prefrontal cortex and the amygdala (35, 36, 39), the activity of which are negatively and positively related to trait anxiety in humans (40, 41). Whereas activation of amygdaloid structures is associated with heightened cardiovascular activity (42, 43), the prefrontal cortex may reduce cardiovascular reactivity to stress (44). Chronic hypoactivity of areas such as the prefrontal cortex, and hyperactivity of limbic regions including the amygdala in individuals with elevated trait anxiety (40, 41) may directly and indirectly interact with hypothalamic and brainstem regions to modulate resting autonomic and cardiovascular activity. Whether these central alterations are responsible for the present relationship between trait anxiety, MSNA, and blood pressure remains unknown.

Leveraging a large data set allowed assessment of subtle interactions between trait anxiety and sympathetic activity that may not have been detectable in previous studies. Furthermore, the current study viewed trait anxiety along a spectrum, as opposed to a dichotomous variable whereby participants are separated based on anxiety disorder diagnosis. This approach is responsive to the need for assessment of psychological variables across complex spectrums within physiological research (45), and offers evidence of a quantifiable relationship between subjective, self-perceived trait anxiety and objective, clinically relevant measures of autonomic and cardiovascular function. The selected approach to trait anxiety evaluation within the current study may also underly discrepant findings between the present study and those performed in clinically diagnosed anxiety disorders (8–12). The accumulation of physiological abnormalities within anxiety and mood disorders has been shown to be impacted by symptom severity in numerous studies (20, 29, 46, 47), supporting the necessity for psychological characteristics to be included as continuous variables within physiological experimentation and analyses.

Although previous studies on participants with anxiety disorders have suggested little to no alterations in multiunit MSNA recordings, considerable evidence supports an underlying difference in sympathetic recruitment patterns (11, 31, 48). Utilization of single-unit MSNA recordings, which allow direct assessment of single vasoconstrictor nerve fibers, have reported a greater probability of single nerve fibers firing, as well as a greater likelihood of single nerve fibers firing multiple times within an integrated burst of MSNA in individuals with PD (11). This augmented MSNA firing pattern has been associated with greater cardiac norepinephrine spillover to plasma (10), indicative of potential regional differences in sympathoexcitation unique to PD that may be shared with general symptoms of anxiety. Similarly, trait anxiety has been shown to be correlated with alterations in the firing pattern of single-unit nerve fibers within individuals with metabolic syndrome (48). It is possible that the healthy adults used in the current study have augmented single-unit firing rates and different sympathetic recruitment strategies as a function of trait anxiety, but we did not have the ability to analyze this given the MSNA data acquisition process within our respective laboratories. Future work investigating sympathetic recruitment patterns may be warranted.

Chronic anxiety and stress are difficult to study within a laboratory for various practical and ethical reasons, making conclusions regarding mechanisms leading to the presently observed relationship difficult to establish. Numerous studies have assessed the cardiovascular and sympathetic response to acute laboratory mental stress in healthy adults to better understand the underlying sympathetic response to psychological stress (13, 49). Mental stressors, such as forced mental arithmetic and the Stroop Color Word task, have been used to evoke a repeatable pressor response (50) across multiple studies (for reviews, see Refs. 13 and 49). Although the pressor response to mental stress has been thought to be partially mediated by sympathetic activity, considerable variations in sympathetic responsiveness have been observed (51), and these changes in sympathetic activity are generally not associated with MAP reactivity (51).

A recent study from our laboratory characterized the sympathetic and cardiovascular response to the trier social stress test (TSST) (52), which uses a period of social stress anticipation devoid of alterations in respiration, muscle movement, and speech before socially evaluated speech and mental arithmetic phases (23). Our findings demonstrated a significant pressor response to anticipatory stress, and a reduction in MSNA in healthy adults (23). These findings are particularly relevant to individuals with chronic anxiety, who often experience general anticipatory stress and anxiety before any stressor presentation. Whether the acute sympathetic and cardiovascular response to the anticipatory stress phase of the TSST observed in our recent study (23) is present across the spectrum of trait anxiety remains unknown. Holwerda et al. (8) reported that individuals with GAD showed augmented MSNA burst amplitude responsiveness during the anticipation of mental arithmetic compared with controls, indicating an exaggerated sympathetic response to anticipatory stress in this population (8). Similarly, Callister et al. (53) observed a positive relationship between perceived stress and MSNA reactivity to mental stress. Individuals who experience chronic anxiety exhibit selective attention and greater sensitivity to threat (54, 55), ultimately impacting stress interpretation and likely sympathetic responsiveness. In line with the theory of allostasis and allostatic overload (56), exaggerated threat perception to everyday events, as well as frequent stress presentation may lead to increases in cardiovascular activity, in turn resulting in chronic activation of the sympathetic nervous system and elevated blood pressure as is observed in the present study.

Whether alterations in sympathetic neural reactivity to laboratory mental stress translate to exaggerated blood pressure reactivity or augmented likelihood of hypertension development remains equivocal. However, Fonkoue et al. (57) reported an augmented MSNA reactivity to mental stress in participants with a family history of hypertension when compared with controls without a family history of hypertension. The translation of laboratory mental stress protocols to real-world applications remains an area that requires further substantiation before definitive conclusions can be drawn. Nevertheless, elevated resting MSNA is highly associated with hypertension development (7), and increases in parallel with blood pressure over time (15). The association between trait anxiety and elevated resting MSNA in the current study may lend insight into baseline sympathoexcitation as a mediating factor between anxiety proneness and hypertension, although longitudinal data are necessary to confirm this.

The current study included sex as a predictor variable within the regression analyses. Anxiety disorders disproportionately impact females (58) and sex differences in neural control of cardiovascular regulation are well documented (59–62). Sex differences with respect to psychological variables and cardiovascular outcomes have been noted. For example, young females with a history of myocardial ischemia have increased susceptibility to mental stress-induced myocardial ischemia when compared with young males (63). Experimental evidence from Greaney et al. (64) has supported an adverse role of psychological factors, namely major depressive disorder, on vascular β-adrenergic receptor function leading to impaired vasodilation, an effect that may be exacerbated with chronic psychosocial stress (65, 66). This is particularly relevant in young females whose β-adrenergic receptor activity opposes the vasoconstrictor effects of norepinephrine released from sympathetic varicosities (60, 61, 67). Our laboratory remains attentive to the impacts of sex and sex hormones on psychological and sympathetic regulation in response to stress (22, 51, 59, 68–72). Unfortunately, sensitivity analysis conducted on the present data set indicated that we did not have adequate statistical power to assess the relationship between trait anxiety, MSNA, and blood pressure stratified between biological sexes (i.e., male vs. female). Nonetheless, this is an area of active investigation within our laboratory and will be explored in future work.

We acknowledge the following limitations in our study. First, we included females with and without hormonal contraceptive use. To minimize confounding factors, females in whom contraceptive use was reported were assessed during the low-hormone phase of contraceptive use. Prior work has demonstrated similar MSNA between low-hormone phase and early follicular phase (73–75). More importantly, a separate assessment including only naturally menstruating females revealed similar results as those reported in the current study (data not shown). Second, brachial blood pressures were taken in the supine position in six participants, while seated blood pressure recordings were taken in the vast majority of the study population (i.e., 82 of 88 participants). Analyses were performed with and without these six participants, and the results were unchanged (data not shown). The current study did not include the state form of the STAI, which measures acute anxiety levels, so we cannot rule out the acute impact of state anxiety on the current findings. However, all participants quietly rested for at least 10 min following MSNA signal acquisition and before any active recording, which allowed for accurate recording of MSNA in a relaxed state. In addition, previous studies that have manipulated state anxiety found no impact on MSNA at rest (76). Since the present study included retrospective data, we were unable to exclude individuals based on diagnosed anxiety disorders. It is possible that some individuals had trait anxiety levels which may indicate an undiagnosed clinical anxiety disorder. The presence of undiagnosed anxiety disorders likely would not impact the interpretation of the findings, and instead may reinforce the necessity for accurate screening, diagnosis, and attentiveness to undiagnosed anxiety disorders in otherwise healthy individuals. Previous studies have indicated that severe major depressive disorder (MDD) is associated with elevated baseline MSNA (47). Conversely, others have shown unchanged MSNA in individuals with MDD compared with healthy controls and observed elevated plasma norepinephrine spillover only in those individuals with MDD and comorbid PD (77). Since anxiety and depressive symptoms often overlap (78), it is possible that individuals with heightened trait anxiety in the current data set additionally had elevated depressive symptoms. Future work disentangling the unique effects of trait anxiety and depression on sympathetic activity is warranted. In addition, while cardiovascular and autonomic medication usage was excluded across all studies, the use of anxiolytics and antidepressants were not assessed in all participants, so we cannot account for their potential impact on resting MSNA (12, 47). In the current data set, all participants self-reported the absence of any known cardiometabolic diseases. However, we did not assess plasma lipid levels, glucose tolerance, and other metabolic parameters that may impact sympathetic outflow at rest and cannot rule out the effects of alterations of these variables in the present study. Finally, the present findings cannot apply causation to the relationship between trait anxiety, MAP, and MSNA. The directionality between these relationships is difficult to assess given various practical and ethical constraints. Anxiety symptoms and the development of anxiety disorders may be exacerbated following acute cardiovascular events (79–81). Though it is unlikely that this relationship exists in the current study population of individuals with no known cardiovascular or autonomic diseases, future prospective investigations following natural changes in trait anxiety across the lifespan may be appropriate and could provide key insights into potential causal relationships between anxiety, blood pressure, and MSNA.

Conclusions

The present study assessed the relationship between trait anxiety, blood pressure, and direct assessment of MSNA in a robust sample of healthy adults. Our findings indicate a direct and independent association between trait anxiety severity, resting blood pressure, and MSNA in healthy adults without comorbid cardiovascular or metabolic disorders. These findings signify an important role for trait anxiety on sympathetic neural control and cardiovascular function at rest and highlight the necessity to consider psychological characteristics when assessing physiological function.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was supported by National Institutes of Health Grants AA0024892, 1R15HL140596-01, AG064038-01A1, U54GM115371, and P20GM103474 and National Center for Advancing Translational Sciences of the National Institutes of Health Award 2TL1 TR002318.

DISCLAIMERS

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

DISCLOSURES

Jason Carter is an editor of American Journal of Physiology-Heart and Circulatory Physiology and was not involved and did not have access to information regarding the peer-review process or final disposition of this article. An alternate editor oversaw the peer-review and decision-making process for this article. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

J.A.B. and J.R.C. conceived and designed research; J.A.B., J.J.D., I.M.G., M.K-R., and J.R.C. performed experiments; J.A.B., J.J.D., I.M.G., M.K-R., and J.R.C. analyzed data; J.A.B., J.J.D., I.M.G., M.K-R., and J.R.C. interpreted results of experiments; J.A.B. prepared figures; J.A.B. drafted manuscript; J.A.B., J.J.D., I.M.G., M.K-R., and J.R.C. edited and revised manuscript; J.A.B., J.J.D., I.M.G., M.K-R., and J.R.C. approved final version of manuscript.