Abstract
Background
Baseline state anxiety, which reflects an individual’s initial anxiety level before stressors, offers an important perspective for understanding how pre-existing anxiety influences responses to subsequent stressors.
Methods
This study investigated physiological changes in 59 participants grouped by low (n = 32, LbSA group) and high (n = 27, HbSA group) baseline state anxiety levels, measured using the Short State Anxiety Inventory. Participants underwent three phases: baseline (T1), the modified Trier Social Stress Test (TSST) task (T2), and recovery (T3). ECG signals were recorded and six heart rate variability (HRV) features were analyzed across time, frequency, and non-linear domains using a mixed linear model.
Results
In T2, the LbSA group showed significantly increased SDNN (t= 2.662, p= 0.027) and decreased RR_Mean (t= -6.819, p< 0.001) and SampEn (t= -2.985, p= 0.010). In contrast, the HbSA group only showed a decreased RR_Mean (t= -5.641, p< 0.001). In T3, the HRV changes in the LbSA group rebounded, whereas the HbSA group did not, resulting in significantly increased SDNN (t= 3.187, p= 0.006), RMSSD (t= 3.719, p= 0.001) and LnLF (t= 3.800, p= 0.001). Additionally, between-group differences reached significance for SampEn in T2 (t= 1.301, p= 0.044) and for SDNN (t= 2.215, p= 0.036) and RMSSD (t= 2.209, p= 0.046) in T3.
Conclusion
Compared to the individuals with low baseline state anxiety, which shows rapid changes under stress and quick recovery afterward, the individuals with high baseline state anxiety exhibit blunted and prolonged responses. The findings highlight the potential influence of baseline state anxiety on autonomic regulatory capacity, suggesting the importance of modulating state anxiety.
Keywords: Baseline state anxiety, Heart rate variability, Stress, Recovery, Trier Social Stress Test
Introduction
Anxiety is an emotion characterized by apprehension and somatic symptoms of tension in which an individual anticipates impending danger, catastrophe, or misfortune. The body often mobilizes itself to meet the perceived threat: Muscles become tense, breathing is faster, and the heart beats more rapidly [1]. Anxiety is currently a significant focus within the field of health psychology, as it is associated with many mental and physical health conditions [2–4]. It is reported that about 34% of adults in the US suffer from anxiety disorders during their lifetime, which are associated with significant distress and impairment [5]. Therefore, numerous studies have been conducted to explore the effects of anxiety, methods of assessment, and intervention techniques, to understand and treat anxiety further.
Anxiety can be categorized into trait and state anxiety [6]. Trait anxiety is a stable personality characteristic, often linked to various psychopathological conditions and chronic high arousal. In contrast, state anxiety is a transient emotional response, characterized by a temporary increase in sympathetic nervous system activity, without specific pathological conditions. Compared with trait anxiety, exploring state anxiety contributes to the short-term and dynamic effects of anxiety, providing valuable insights for assessment and intervention.
State anxiety is commonly measured using self-report scales like the state-anxiety subscale of Spielberger State Anxiety Scale (SSAI; [6]). However, existing measures do not adequately assess all components of anxiety [7] due to their subjectivity and lack of real-time assessment capabilities. Physiological measures, particularly heart rate variability (HRV), offer valuable physiological information. As a physiological measure of autonomic nervous system (ANS) activity, HRV provides non-invasive, real-time information on heart rate modulation by the ANS [8], indicating the balance between the sympathetic (i.e., “fight, flight, or freeze”) and parasympathetic (i.e., “rest and digest”) control of the heart [9]. Different HRV indices reflect distinct patterns of parasympathetic and sympathetic control over the nervous system [9], which are associated with varying capacities for self-regulation and environmental adaptation. HRV’s sensitivity makes it a robust biomarker for dynamically assessing anxiety [10].
Acute stress serves as a potent trigger for state anxiety. When exposed to external stressors (e.g., social evaluation), individuals experience rapid increases in anxiety levels alongside physiological arousal [11]. The Trier Social Stress Test (TSST) capitalizes on this mechanism by standardizing a social-evaluative threat to reliably induce state anxiety [12], serves as an ideal paradigm for investigating the relationship between state anxiety and HRV. During TSST-induced stress, individuals exhibit not only a rapid increase in stress and state anxiety [13], but also significant activation of the ANS, manifested as rapid changes in HRV [14–16]. However, existing research has predominantly classified study populations based on trait anxiety or comparisons between anxiety patients and healthy controls when examining HRV responses to acute stress. Notably, these studies have yielded inconsistent findings: while some studies demonstrate that individuals with high trait anxiety or anxiety disorders exhibit lower high-frequency HRV compared to those with low trait anxiety under stress [17], others report non-significant group differences [18, 19]. These discrepancies suggest that trait anxiety- as a stable, enduring psychological characteristic - may inadequately reflect dynamic physiological responses to acute stress, resulting in a temporal mismatch with actual psychophysiological states during acute stress tasks.
In contrast, baseline state anxiety levels may more directly reflect an individual’s initial psychophysiological state when facing an impending stressor, potentially representing an unexamined transient influence in prior research. However, existing studies have primarily focused on how anticipatory mental state modulates HRV responses to acute stress, as exemplified by Zhao et al.’s [20] study of shooting athletes where first-class athletes with lower pre-competition state anxiety demonstrated superior emotion regulation and higher HRV during competition. This predominant focus on anticipatory contexts has created a critical gap in understanding how non-anticipatory baseline state anxiety - measured without explicit stressor foreknowledge - regulates physiological responses. Whereas anticipatory mental state involves conscious preparation for known challenges [21], baseline state anxiety reflects spontaneous autonomic fluctuations independent of top-down expectation processes. Conducting explicit pre-grouping based on baseline state anxiety levels and performing phase-specific comparisons (including baseline, stress task, and recovery) could help disentangle the effects of baseline anxiety states from active stress anticipation, elucidate its independent role in HRV dynamics, and provide a more comprehensive reflection of how state anxiety influences individuals’ autonomic regulation capabilities, informing a basis for clinical or targeted interventions.
Therefore, this study aims to investigate how HRV changes in individuals with high versus low baseline state anxiety across three distinct phases: baseline, the TSST task phase, and recovery phase. We propose the following exploratory hypotheses: (1) There are differences in HRV at baseline between the high baseline state anxiety group and low baseline state anxiety group; (2) The pattern of HRV responses under stress varies between the two groups; (3) Significant between-group differences are anticipated in the recovery phase.
Methods
Participants
Based on an effect size (d= 0.84) reported by Held et al. [19] and calculated via G*Power 3.1 (α = 0.05, power = 0.80, two-sample t-test), a minimum sample of N = 50 was required. To account for potential issues such as poor signal quality when using physiological equipment, sixty-six participants were recruited from the university through convenience sampling. All subjects confirmed that they had no history of (1) mental illness or use of psychiatric medications, (2) cardiovascular diseases, or use of any medications known to affect cardiovascular measurements. Seven subjects were excluded from the study: 3 for premature withdrawal, 2 for improper operations, and 2 for excessive noise and poor quality of ECG signals. Consequently, a total of 59 valid participants (Female: 26, Male: 33; Age = 22.29 ± 3.05 years) were included in the analysis - exceeding the required sample size and providing >80% statistical power.
This study was conducted in accordance with the Declaration of Helsinki and approved by the Medical Ethics Committee, Department of Psychological and Behavioral Sciences, Zhejiang University, China (Approval No. 2023034). All participants were carefully informed about the study protocol and written informed consent was obtained prior to the experiment. The consent form described the study as a“social scenario experiment” rather than interview to ensure the measurement of baseline data in a calm state without prematurely inducing anxiety.
Experimental scenario and procedure
The experimental scenario, shown in Fig.1, was conducted in a closed laboratory following TSST Paradigm guide proposed by Labuschagne et al. [22], with the participant and experimenters seated facing each other. A laptop was provided for the subject to access instructions, complete psychological assessments, and signal the experiment’s stage with a tone. Two experimenters, dressed in formal attire and seated 1.5 meters away, held scoring sheets and interview instructions. A 500W light source and a Nikon digital single-lens reflex (DSLR) camera, positioned 1 meter away, were used to enhance self-focus by drawing the participant’s attention to their performance.
Fig. 1.
Experimental scenario
Participants were briefed on prerequisites for physiological signal reliability: maintaining a regular diet, abstaining from stimulants like tobacco, alcohol, tea, and coffee, and ensuring over 6 h of sleep the night before the experiment.
The experimental session began with a preparation phase where participants were guided through ECG sensor placement and completed demographic questionnaires and psychological scales in the laboratory (with interviewers waiting nearby and environmental equipment not yet activated). The formal procedure consisted of three phases: baseline phase (T1), task phase (T2), and recovery phase (T3), with questionnaires completed after each to assess psychological state (see Fig. 2 for details). During baseline period, the participants were instructed to sit in a relaxed and comfortable position. Afterward, participants entered a 10-minute task phase (2-minute task-introduction, 5-minute English task and 3-minute calculation task), starting with a self-introduction and task explanation under spotlight and camera. They were tasked with an English presentation on “My career aspirations” followed by an interview focusing on strengths and weaknesses related to their career goals. This was succeeded by a calculation task, requiring participants to subtract 17 from 2024 repeatedly for three minutes, correcting any mistakes. Finally, a five-minute recovery period followed for emotional regulation.
Fig. 2.
Experimental procedure. SIAS, social interaction anxiety scale. SSAI, short state anxiety inventory
Questionnaires
Social Interaction Anxiety Scale
The Social Interaction Anxiety Scale (SIAS) was developed by Mattick [23] and revised by Ye [24]. It has 19 items with a 5-point scale (1~5 for“very non-compliant” to “very compliant”). Questions 8 and 10 are reverse scored. The total score of the SIAS is 95 points and the higher the score, the higher the level of social anxiety. The Cronbach’s alpha coefficient for the internal consistency reliability of the Chinese-version SIAS was 0.874 [24]. Participants were asked to complete SIAS during the preparation phase.
Short State Anxiety Inventory
The Short State Anxiety Inventory (SSAI) [25], as revised by Tian et al. [26], was used to assess the subject’s immediate fears, nervousness, worries, and neurotic experiences or feelings. This scale consists of 6 items with 2 factors of anxiety presence and anxiety absence. Each item is scored on a 4-point scale from 1 (not at all) to 4 (almost always), with the anxiety absence items reverse scored. The total score of the SSAI is 24 points and higher scores indicate higher levels of anxiety. The Cronbach’s alpha coefficient for the internal consistency reliability of the Chinese-version scale is 0.82 [26]. Participants completed SSAI after each phase to assess the current psychological state.
ECG Data Recording and Processing
Throughout the experiment, a dual-electrode, single-channel portable electrocardiographic device (Holter monitor, CP-02x model, Beneware, Hangzhou, China) with a sampling frequency of 200 Hz was used to continuously monitor ECG signals. The positive electrode (+) was positioned at the lower end of the sternum, while the negative electrode (-) was placed at the upper end of the sternum. The raw ECG data were processed using the accompanying HolterSystem software to generate the initial R-peak position sequence. Misidentified R-peaks caused by artifacts and missing R-peaks were manually corrected by visually inspecting all ECG signals, with the assistance of tools provided by the software.
The R-peak position sequence extracted from the software was exported for offline analysis. For each subject, the sequence was divided into three distinct phases according to the experimental protocol and RR interval series were computed for each phase: a baseline phase of 300 s, a task phase of 480 s (analyzing only the last 8 minutes of RR intervals after excluding the initial 2-minute preparation period), and a recovery phase of 300 s.
Six HRV features (shown in Table 1) were computed for each phase across time, frequency, and nonlinear domains using NeuroKit2 [27], an open-source Python toolbox for physiological signal processing. These carefully selected metrics are recommended features for characterizing autonomic activity and are also very commonly used in anxiety research [28, 29].
Table 1.
HRV features utilized in the present study
| Features (unit) | Description | Biological interpretation |
|---|---|---|
| RR_Mean (ms) | Mean of RR intervals | Negative correlated to heart rate |
| SDNN (ms) | Standard deviation of all normal-to-normal (NN) intervals | Represents short-term variability and both SNS and the parasympathetic nervous system (PNS activity contribute to SDNN |
| RMSSD (ms) | Root mean square of successive NN interval differences | Influenced by the PNS |
| LnLF | Natural logarithm of power in low frequency range of 0.04-0.15 Hz | May be produced by both the PNS and SNS |
| LnHF | Natural logarithm of power in high frequency range of 0.15-0.4 Hz | Correlated with RMSSD. Reflects PNS activity |
| SampEn | Sample entropy as a measure of complexity | Reflect the complexity of signal |
Data analysis
Participants were categorized into two groups based on their baseline SSAI scale scores (SSAI-T1 scores) using a median split (median= 10): high baseline state anxiety group (HbSA Group) with 27 subjects scoring strictly > 10, and low baseline state anxiety group (LbSA Group) with 32 subjects scoring ≤ 10. The median split approach [30, 31] was justified to detect nonlinear discontinuities in stress reactivity/recovery across different baseline state anxiety given the absence of clinical cutoffs.
Initially, the student’s t-test was applied to compare the group differences in demographic characteristics, including age, SIAS scores, and SSAI-T1 scores. Specifically, the difference in gender was examined using chi-square test. Meanwhile, the impact of covariates, namely gender and age, on HRV features was evaluated using Linear Mixed Model (LMM) analysis. To explore the effects of time and group, a mixed linear model was conducted to assess the changes in HRV features from the baseline to the recovery phase within each group and the between-group differences in HRV features during T1, T2 and T3. The statistical model employed was as follows:
. In this model, HRV represented the specific HRV features as the dependent variable, Group (HbSA or LbSA group) and Time (T1, T2, T3), along with their interaction, were treated as fixed effects. Each participant (id) was incorporated with random intercepts. Parameter estimation was conducted using the restricted maximum likelihood (REML) method, and degrees of freedom were calculated using the Satterthwaite method. All statistical analyses were performed using IBM SPSS Statistics 26. We performed two-tailed t-tests for demographic analyses and employed LMM for statistical modeling. To account for multiple comparisons, we applied Bonferroni correction, with statistical significance determined by adjusted p-values (hereafter denoted as “p”)< 0.05. Effect sizes were estimated using Hedges’ g (denoted as g) with 95% confidence intervals.
Results
Demographic characteristics
Fifty-nine participants were available for analysis in the study. No significant differences were observed in sex distribution (p= 0.317) or SIAS scores (p= 0.456) between the two groups. However, there were notable differences in age and baseline SSAI scores. The HbSA group were statistically younger (M= 20.7, SD= 2.2; t (57) = 3.980, p< 0.001) and had higher baseline SSAI scores (M= 12.8, SD= 2.1; t (57) = -9.439, p< 0.001) compared to the LbSA group (age: M= 23.6, SD= 3.1; SSAI: M= 8.3, SD= 1.5). Group characteristics are provided in Table 2.
Table 2.
Demographic characteristics of the HbSA and LbSA groups
| HbSA Group (n = 27) | LbSA Group (n = 32) | df | t/
|
P-value | |
|---|---|---|---|---|---|
| Sex (male/female) | 10/17 | 16/16 | 1 | 1.00 | 0.317 |
| Age (years) | 20.7 ± 2.2 | 23.6 ± 3.1 | 57 | 3.980 | < 0.001 |
| SIAS | 49.7 ± 12.0 | 47.3 ± 11.5 | 55 | −0.751 | 0.456 |
| SSAI-T1 | 12.8 ± 2.1 | 8.3 ± 1.5 | 57 | −9.439 | < 0.001 |
Data are expressed as mean ± standard deviation . SIAS social interaction anxiety scale, SSAI short state anxiety inventory
The GLM analysis, controlling for age and sex as covariates, revealed that neither sex (p= 0.313) nor age (p= 0.408) had a significant effect on HRV features. Since including non-significant covariates may reduce statistical power in mixed-effects models by unnecessarily increasing model complexity, we therefore excluded age and sex as covariates in the final linear mixed model’s analysis.
Changes in SSAI scores and HRV features across experimental conditions
Following the stressful tasks, SSAI scores significantly increased in both groups (LbSA: t (112) = 10.856, p< 0.001, g= 1.72; HbSA: t (112) = 4.010, p< 0.001, g= 0.84), indicating that state anxiety was induced by the tasks. During the recovery phase, the scores remained elevated, without a notable reduction (LbSA: t (112) = -0.002, p= 1.000, g= -0.002; HbSA: t (114) = -1.771, p= 0.237, g= 0.36).
In terms of time-domain measures, both groups initially showed a substantial decrease in RR_mean during the stress phase (LbSA: t (113) = -6.819, p< 0.001, g= -0.74; HbSA: t (113) = -5.641, p< 0.001, g= -0.82), which was succeeded by a pronounced rebound in recovery (LbSA: t (113) = 6.701, p< 0.001, g= 0.73; HbSA: t (113) = 7.332, p< 0.001, g= 1.08), with the HbSA group demonstrating a robust recovery that exceeded their baseline levels (t (113) = 1.767, p= 0.240, g= 0.27). SDNN in the LbSA group responded to stress with a significant increase (t (113) = 2.662, p= 0.027, g= 0.44), whereas the HbSA group displayed an increase relative to the baseline only during the recovery phase (t (113) = 3.187, p= 0.006, g= 0.58). RMSSD in the LbSA group was not significant during both the task phase and the recovery phase. In contrast, the HbSA group displayed a significant rise in the recovery phase (t (113) = 3.058, p= 0.008, g= 0.41), ultimately exceeding baseline levels significantly (t (113) = 3.719, p= 0.001, g= 0.49).
Regarding frequency-domain measures, the LbSA group showed no significant fluctuations in LnLF (T2-T1: t (113) = 2.274, p= 0.077, g= 0.35; T3-T2: t (113) = -0.442, p= 1.000, g= 0.08) and LnHF (T2-T1: t (113) = 0.072, p= 1.000, g= 0.01; T3-T2: t (113) = 1.476, p= 0.429, g= 0.23) but showed significant elevation at T3 relative to T1 in LnLF (t (113) = 2.717, p= 0.024, g= 0.43). In the HbSA group, LnLF (t (113) = 1.072, p= 0.850, g= 0.22) and LnHF (t (113) = 0.551, p= 0.586, g= 0.02) also showed no significant fluctuations during stress but surged in the recovery phase (LnLF: t (113) = 2.736, p= 0.022, g= 0.61; LnHF: t (113) = 2.109, p= 0.109, g= 0.33), surpassing pre-stress levels (LF: t (113) = 3.800, p= 0.001, g= 0.82; HF: t (113) = 2.306, p= 0.068, g= 0.35).
As for the nonlinear measure of SampEn, stress led to a significant drop only in the LbSA group (t (113) = -2.985, p= 0.010, g= -0.64), whereas there was no change in the HbSA group.
Data are presented in Table 3 for clarity.
Table 3.
SSAI scores and HRV features in HbSA and LbSA groups in T1, T2 and T3
| HbSA (n = 27) | LbSA(n = 32) | InteractionF (p) Group:Time | Simple Effect p | Hedges’s g with 95% CI |
|
|---|---|---|---|---|---|
| SSAI | 10.16 (<0.001) | ||||
| T1 | 12.81 ± 2.15 | 8.34±1.47 | <0.001 | 2.42 [1.67, 3.17] | |
| T2 | 16.32 ± 4.86 * | 17.17±4.81* | 0.400 | −0.18 [−0.75,0.39] | |
| T3 | 17.96 ± 2.84 * | 17.16±3.32* | 0.377 | 0.26 [−0.31, 0.83] | |
| RR_mean | 1.00 (0.371) | ||||
| T1 | 756.12 ± 95.01 | 768.03±139.74 | 0.701 | −0.10 [−0.67, 0.47] | |
| T2 | 677.05 ± 95.93 * | 680.25 ± 111.83 * | 0.918 | −0.03 [−0.60, 0.54] | |
| T3 | 782.16 ± 104.84 # | 766.52 ± 145.31 # | 0.637 | 0.13 [−0.44,0.70] | |
| SDNN | 2.139 (0.122) | ||||
| T1 | 60.39 ± 25.30 | 53.29 ± 19.11 | 0.222 | 0.31 [−0.26, 0.88] | |
| T2 | 64.24 ± 21.00 | 62.05 ± 21.33 * | 0.705 | 0.10 [−0.47, 0.67] | |
| T3 | 72.59 ± 28.80 * | 59.63 ± 16.68 | 0.036 | 0.57 [0.00, 1.14] | |
| RMSSD | 2.558 (0.082) | ||||
| T1 | 37.05 ± 17.58 | 34.48 ± 16.92 | 0.576 | 0.15 [−0.42, 0.72] | |
| T2 | 38.52 ± 17.53 | 32.03±13.38 | 0.161 | 0.41 [−0.16, 0.98] | |
| T3 | 45.58±24.61*# | 35.99±14.80 | 0.046 | 0.51 [−0.06, 1.08] | |
| LnLF | 1.497 (0.228) | ||||
| T1 | 6.67 ± 0.86 | 6.51 ± 0.80 | 0.422 | 0.19 [−0.38, 0.76] | |
| T2 | 6.80 ± 0.59 | 6.76 ± 0.73 | 0.860 | 0.06 [−0.51, 0.63] | |
| T3 | 7.16 ± 0.82 *# | 6.82 ± 0.69 * | 0.101 | 0.45 [−0.12, 1.02] | |
| LnHF | 0.264 (0.769) | ||||
| T1 | 6.10 ± 0.97 | 5.80 ± 1.28 | 0.262 | 0.27 [−0.30, 0.84] | |
| T2 | 6.12 ± 0.85 | 5.81 ± 0.90 | 0.235 | 0.35 [−0.22, 0.92] | |
| T3 | 6.44 ± 1.00 | 6.02 ± 1.00 | 0.115 | 0.42 [−0.15, 0.99] | |
| SampEn | 1.877 (0.158) | ||||
| T1 | 1.41 ± 0.30 | 1.43±0.27 | 0.803 | −0.07 [−0.64, 0.50] | |
| T2 | 1.39±0.30 | 1.23±0.34* | 0.044 | 0.52 [−0.05, 1.09] | |
| T3 | 1.40 ± 0.33 | 1.31±0.34* | 0.191 | 0.27 [−0.30,0.84] |
Data are presented as mean ± standard deviation; * indicates p < 0.05 compared to T1 within the same group; # indicates p < 0.05 compared to T2 within the same group. Interaction F(p) indicates F-statistic and p-value for Group*Time interaction from mixed model and Simple Effect p indicates p-values for between-group comparisons at each phase.p-values for multiple comparisons were adjusted using the Bonferroni method
Differences in SSAI scores and HRV features between HbSA and LbSA groups
As shown in Table 3, significant differences in SSAI scores between the two groups were observed at baseline (t (160) = 4.968, p< 0.001), but the discrepancy disappeared after the stress was introduced (T2: t (161) = -0.844, p= 0.400; T3: t (160) = -0.864, p= 0.377) (Fig.3). Concerning HRV features, only SampEn (t (149) = 1.301, p= 0.044, g= 0.52) showed significant between-group differences in T2, SDNN (t (94) = 2.125, p= 0.036, g= 0.57) and RMSSD (t (77) = 2.029, p= 0.046, g= 0.51) revealed statistical between-group differences specifically in T3 (Fig.4).
Fig. 3.
SSAI scores in HbSA and LbSA groups during baseline, task, and recovery
Fig. 4.
HRV features in HbSA and LbSA groups during baseline, task, and recovery. Horizontal significance markers label significant changes in phases within groups. Vertical significance markers label significant between-group differences
Discussion
In this study, we aimed to explore how baseline state anxiety levels influence individuals’ self-regulation abilities under stress. To provide insights into the underlying physiological dynamics, the changes in HRV across different phases, including the resting state, stress state, and recovery phase, were analyzed.
In Hypothesis 1, we investigated the resting HRV differences between individuals with high and low baseline state anxiety. Contrary to our expectations, no significant between-group differences were observed in HRV. We suggest that this lack of significance may be primarily due to the fact that the HbSA group in our study had SSAI scores that were relatively high (average of 12.81) but not severe enough. On one hand, previous studies with statistical results on baseline HRV typically compare individuals with anxiety disorders to healthy controls [17, 32]. On the other hand, the findings are not consistent across all studies, with some studies failing to find such significant differences [33, 34]. These mixed findings also suggest that the baseline state may not provide sufficient information to discern differences in autonomic regulatory abilities between groups. Instead, focusing on the physiological changes during dynamic phases, such as the stress and recovery phases, may offer greater value in understanding the influence of baseline state anxiety.
Therefore, in Hypothesis 2, we assessed the stress responses between individuals with high and low baseline state anxiety. Consistent with our hypothesis, there were differences in the HRV change patterns between the two groups under stress. Individuals with low baseline state anxiety showed a more rapid and pronounced response to stress, whereas those with high baseline state anxiety exhibited a blunted change in HRV. Specifically, individuals with low baseline state anxiety showed sympathetic activation under stress, with significant decreases in RR_Mean (i.e., an increase in heart rate) and SampEn (i.e., an increase in regularity, this decrease has been found in many sympathetically dominated activities like REM [35]). These findings are in concordance with previous research [16, 36] and this phenomenon aligns with the fight-or-flight response theory, which support the idea that there was an inhibition of the parasympathetic system and a prevalence of the sympathetic system during a stressing situation [37]. Compared to the significant response of individuals with low baseline state anxiety to stress, individuals with high baseline state anxiety seem to exhibit a blunted HRV response, with only heart rate showing a significant increase, and no significant changes in other features. This blunted stress response corresponds with the findings of previous research, which have reported that the HF-HRV response to mild stress is blunted in those with anxiety disorders [38]. In addition, the anxious individuals may exhibit reduced reactivity of their defense system under chronic stress, long-term anxiety experiences, and the potential development of“learned helplessness”. This reduction can lead to blunted responses when facing threats, as previously observed in autonomic measures [39] and diminished defensive responses [40]. Such findings may, in part, explain why individuals with high baseline state anxiety do not exhibit responses to stress that are as pronounced as those observed in individuals with low baseline state anxiety.
Additionally, it is interesting to note that there were no differences in heart rate changes between the two groups, suggesting that baseline state anxiety levels may not influence the degree of heart rate increase under stress, which is also observed in the studies by Trotman [41]. This highlights the limitations of average heart rate changes as an indicator of stress responses and underscores the need to focus on more nuanced features derived from the cardiac signal, such as HRV, which are more sensitive and informative, to gain a more comprehensive understanding of individuals’ physiological and psychological reactions under stress.
In Hypothesis 3, we examined the HRV responses during the recovery phase between individuals with high and low baseline state anxiety. Consistent with the hypothesis, significant differences during the recovery phase were found between the two groups. Compared to the recovery trend of individuals with low baseline state anxiety, those with high baseline state anxiety showed prolonged stress responses, and the differences between the two groups became significant. Although the SSAI scale indicates that individuals with low baseline state anxiety have not fully recovered, most of their HRV features have basically returned to baseline while LF power remained significantly elevated, meaning that individuals with low baseline state anxiety have relatively rapid responses and recoveries to stress; whereas individuals with high baseline state anxiety showed significant differences from baseline in most HRV features, and resting did not restore their stress state; their stress responses prolonged after the stress had ended. Efficient recovery is paramount for maintaining physiological homeostasis and preventing the cumulative wear-and-tear associated with chronic stress [42]. Our findings suggest that individuals entering a situation with higher baseline state anxiety may be less equipped to “bounce back” physiologically from acute stressors they encounter. This impaired recovery capacity is also in line with the sustained cognitive hypothesis [43], which emphasizes the prolonging effect of anxiety characteristics, particularly those of generalized anxiety disorder, on stress-related physiological and emotional activation. Measuring baseline state anxiety levels prior to a stressor might predict the magnitude and duration of this physiological perseveration, specifically indexed by HRV recovery.
These physiological recovery differences following stress, attributable to varying baseline state anxiety, highlight the potential difficulties in stress recovery for individuals with high baseline state anxiety, providing new insights into the complexity of stress responses. In addition, the dissociation between subjective report (SSAI indicating incomplete recovery in both groups) and objective HRV measures (near-baseline recovery in the LbSA group vs. persistent deviation in the HbSA group) also suggests the importance of objective parameter measures.
However, several limitations of the current study are worth noting. Firstly, the difference in the average SSAI scores between our two groups of participants was relatively small (the HbSA group had an average score of 12.8, whereas the LbSA group had an average score of 8.3). Including individuals with higher baseline SSAI scores or greater stress reactivity would have enhanced the sensitivity and specificity of the study results. Secondly, the stressor used in our study was conducted in a laboratory setting, which differs from the stress situations people face in real life. Future research could consider expanding to real-life stress situations, focusing on individuals’ stress coping behaviors and physiological changes in everyday contexts, thereby increasing the ecological validity and practical significance of the study results. Finally, our HRV models did not account for potential confounding variables such as BMI and past training history. Future studies should incorporate these covariates to improve the robustness and generalizability of the findings.
In summary, our findings reveal no discernible in baseline HRV between individuals with high and low baseline state anxiety, but we did observe divergent patterns in their stress responses and subsequent recovery to stress. Specifically, Individuals with low baseline state anxiety generally show rapid and significant increases in sympathetic nervous activity under stress, followed by quick recovery afterward; whereas individuals with high baseline state anxiety show blunted and prolonged responses. This suggests the potential influence of state anxiety on autonomic regulatory capacity and highlights the importance of modulating state anxiety.
Acknowledgements
Special thanks to Beneware Company Hangzhou China for sponsoring the HRV device. We also thank Huiling Zhou for her assistance in data collection, and all participants involved in this study.
Authors’ contributions
YX: Conceptualization, Formal analysis, Validation, Visualization, Writing – original draft, Writing – review & editing. YW: Conceptualization, Formal analysis, Validation, Writing – original draft, Writing – review & editing. MX: Data curation, Writing – original draft, Writing – review & editing. HZ: Data curation, Investigation, Methodology. JZ: Writing – review & editing.HC: Supervision, Writing – review & editing. SC: Project administration, Resources, Writing – review & editing. WC: Conceptualization, Methodology, Project administration, Writing – review & editing.
Funding
This research was supported by Scientific Research Foundation of Hang Zhou City University (No. QD2025017).
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
This study was conducted in accordance with the Declaration of Helsinki and approved by the Medical Ethics Committee, Department of Psychological and Behavioral Sciences, Zhejiang University, China (Approval No. 2023034). All participants were carefully informed about the study protocol and written informed consent was obtained prior to the experiment.
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1. American Psychological Association. APA dictionary of psychology (2nd ed.). American Psychological Association. 2020
- 2.Bisgaard TH, Allin KH, Keefer L, Ananthakrishnan AN, Jess T. Depression and anxiety in inflammatory bowel disease: epidemiology, mechanisms and treatment. Nature Reviews Gastroenterology & Hepatology. 2022;19(11):717–26. 10.1038/s41575-022-00634-6. [DOI] [PubMed] [Google Scholar]
- 3.Chen C. Recent advances in the study of the comorbidity of depressive and anxiety disorders. Adv Clin Exp Med. 2022;31(4):355–8. 10.17219/acem/147441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mendez MF. The relationship between anxiety and Alzheimer’s disease. J Alzheimer’s Disease Rep. 2021;5(1):171–7. 10.3233/ADR-210294. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Szuhany KL, Simon NM. Anxiety disorders: a review. JAMA. 2022;328(24):2431–45. 10.1001/jama.2022.2274. [DOI] [PubMed] [Google Scholar]
- 6. Spielberger CD, Gonzalez-Reigosa F, Martinez-Urrutia A, Natalicio LF, Natalicio DS. The State-Trait Anxiety Inventory. Interam J Psychol. 1971;5(3&4).
- 7. Lg W. The measurement of anxiety. Postgraduate Medical Journal. 1990;66(Suppl 2):S11-7. [PubMed]
- 8.Pham T, Lau ZJ, Chen SHA, Makowski D. Heart rate variability in psychology: a review of HRV indices and an analysis tutorial. Sensors. 2021;21(12):1–20. 10.3390/s21123998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Heiss S, Vaschillo B, Vaschillo EG, Timko CA, Hormes JM. Heart rate variability as a biobehavioral marker of diverse psychopathologies: a review and argument for an ideal range. Neurosci Biobehav Rev. 2021;121:144–55. 10.1016/j.neubiorev.2020.12.004. [DOI] [PubMed] [Google Scholar]
- 10.Porges SW. The polyvagal perspective. Biol Psychol. 2007;74(2):116–43. 10.1016/j.biopsycho.2006.06.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Schneiderman N, Ironson G, Siegel SD. Stress and health: psychological, behavioral, and biological determinants. Annu Rev Clin Psychol. 2005;1:607–28. 10.1146/annurev.clinpsy.1.102803.144141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kirschbaum C, Pirke KM, Hellhammer DH. The ‘Trier Social Stress Test’--A tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology. 1993; 28(1-2):76–81. 10.1159/000119004. [DOI] [PubMed]
- 13.Allen AP, Kennedy PJ, Cryan JF, Dinan TG, Clarke G. Biological and psychological markers of stress in humans: focus on the Trier social stress test. Neurosci Biobehav Rev. 2014;38:94–124. [DOI] [PubMed] [Google Scholar]
- 14.Thayer JF, Ahs F, Fredrikson M, Sollers J. J. III, Wager TD. A meta-analysis of heart rate variability and neuroimaging studies: implications for heart rate variability as a marker of stress and health. Neurosci Biobehav Rev. 2012;36(2):747–56. [DOI] [PubMed] [Google Scholar]
- 15.Cheah C, Lavery C, Johnson AR, Clarke PJF, Hyett MP, McEvoy PM. Changes and persistence in heart rate variability before and during social stress: a comparison of individuals with and without social anxiety disorder. J Anxiety Disord. 2025;110:102960. 10.1016/j.janxdis.2024.102960. [DOI] [PubMed] [Google Scholar]
- 16.Castaldo R, Melillo P, Bracale U, Caserta M, Triassi M, Pecchia L. Acute mental stress assessment via short term HRV analysis in healthy adults: a systematic review with meta-analysis. Biomed Signal Process Control. 2015;18:370–7. 10.1016/j.bspc.2015.02.012. [Google Scholar]
- 17.Beauchaine TP, Thayer JF. Heart rate variability as a transdiagnostic biomarker of psychopathology. Int J Psychophysiol. 2015;98(2 Pt 2):338–50. 10.1016/j.ijpsycho.2015.08.004. [DOI] [PubMed] [Google Scholar]
- 18.Peng HN, Wu JH, Sun XF, Guan Q, Luo YJ. Trait anxiety predicts the response to acute psychological stress. Acta Psychol Sin. 2018;50(9):997–1006. 10.3724/SP.J.1041.2018.00997. [Google Scholar]
- 19.Held J, Vîslă A, Wolfer C, et al. Heart rate variability change during a stressful cognitive task in individuals with anxiety and control participants. BMC Psychol. 2021;9:44. 10.1186/s40359-021-00551-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zhao C, Wang K, Li D, Li Y, Wang Z, Liu Y, Zhang T. Relationship between state anxiety, heart rate variability, and shooting performance in adolescent shooters. BMC Psychol. 2024;12(1):736. 10.1186/s40359-024-02062-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Pulopulos MM, Vanderhasselt M-A, De Raedt R. Association between changes in heart rate variability during the anticipation of a stressful situation and the stress-induced cortisol response. Psychoneuroendocrinology. 2020;114: 104605. 10.1016/j.psyneuen.2020.104605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Labuschagne, I. An introductory guide to conducting the Trier Social Stress Test. Neuroscience and Biobehavioral Reviews. 2019 [DOI] [PubMed] [Google Scholar]
- 23.Mattick RP, Clarke JC. Development and validation of measures of social phobia scrutiny fear and social interaction anxiety. Behav Res Ther. 1998;36(4):455–70. 10.1016/S0005-7967(97)10031-6. [DOI] [PubMed] [Google Scholar]
- 24.Ye D, Qian M-Y, Liu X, Chen X. Revision of social interaction anxiety scale and social phobia scale. Chin J Clin Psychol. 2007;15(2):115–7. [Google Scholar]
- 25.Marteau TM, Bekker H. The development of a six-item short-form of the state scale of the Spielberger State-Trait anxiety inventory (STAI). Br J Clin Psychol. 1992;31(3):301–6. 10.1111/j.2044-8260.1992.tb00997.x. [DOI] [PubMed] [Google Scholar]
- 26.Tian Y, Yang D, Ding C, Cao M. Validity and reliability of the short state anxiety inventory in college students. Chin Mental Health J. 2018;32(10):886–8. [Google Scholar]
- 27.Makowski D, Pham T, Lau ZJ, Brammer JC, Lespinasse F, Pham H, et al. NeuroKit2: a Python toolbox for neurophysiological signal processing. Behav Res Methods. 2021;53(4):1689–96. 10.3758/s13428-020-01516-y. [DOI] [PubMed] [Google Scholar]
- 28.Shaffer F, Ginsberg JP. An overview of heart rate variability metrics and norms. Front Public Health. 2017;5:258. 10.3389/fpubh.2017.00258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kreibig SD. Autonomic nervous system activity in emotion: a review. Biol Psychol. 2010;84(3):394–421. 10.1016/j.biopsycho.2010.03.010. [DOI] [PubMed] [Google Scholar]
- 30.Iacobucci D, Posavac SS, Kardes FR, Schneider MJ, Popovich DL. Toward a more nuanced understanding of the statistical properties of a median split. Psychol Methods. 2015;20(1):102–20. 10.1037/met0000039.25822208 [Google Scholar]
- 31.Iacobucci D, Posavac SS, Kardes FR, Schneider MJ, Popovich DL. The median split: robust, refined, and revived. J Consum Psychol. 2015;25(4):690–704. 10.1016/j.jcps.2015.02.005. [Google Scholar]
- 32.Chalmers JA, Quintana DS, Abbott MJ-A, Kemp AH. Anxiety disorders are associated with reduced heart rate variability: a meta-analysis. Front Psychiatry. 2014;5:80. 10.3389/fpsyt.2014.00080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Chalmers JA, Heathers JAJ, Abbott MJ, Kemp AH, Quintana DS. Worry is associated with robust reductions in heart rate variability: a transdiagnostic study of anxiety psychopathology. BMC Psychol. 2016;4(1):32. 10.1186/s40359-016-0138-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hammel JC, Smitherman TA, McGlynn FD, Mulfinger AM, Lazarte AA, Gothard KD. Vagal influence during worry and cognitive challenge. Anxiety, Stress, & Coping. 2011;24(2):121–36. 10.1080/10615806.2010.490912. [DOI] [PubMed] [Google Scholar]
- 35.Liang X, Xiong J, Cao Z, Wang X, Li J, Liu C. Decreased sample entropy during sleep-to-wake transition in sleep apnea patients. Physiol Meas. 2021. 10.1088/1361-6579/abf1b2. [DOI] [PubMed] [Google Scholar]
- 36.Kim H-G, Cheon E-J, Bai D-S, Lee YH, Koo B-H. Stress and heart rate variability: a meta-analysis and review of the literature. Psychiatry Investig. 2018;15(3):235–45. 10.30773/pi.2017.08.17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Bracha HS. Freeze, flight, fight, fright, faint: Adaptationist perspectives on the acute stress response spectrum. CNS Spectrums. 2004;9(9):679–85. 10.1017/s1092852900001954. [DOI] [PubMed] [Google Scholar]
- 38. Tolin DF. Psychophysiological assessment of stress reactivity and recovery in anxiety disorders. J Anxiety Disord. 2021;82. 10.1016/j.janxdis.2021.102426. [DOI] [PubMed]
- 39.Cook EW, Melamed BG, Cuthbert BN, McNeil DW, Lang PJ. Emotional imagery and the differential diagnosis of anxiety. J Consult Clin Psychol. 1988;56(5):734–40. 10.1037/0022-006X.56.5.734. [DOI] [PubMed] [Google Scholar]
- 40.Lang PJ, McTeague LM. The anxiety disorder spectrum: fear imagery, physiological reactivity, and differential diagnosis. Anxiety Stress Coping. 2009;22(1):5–25. 10.1080/10615800802478247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Trotman GP, Ginty AT, Williams SE. Associations between heart rate, perceived heart rate, and anxiety during acute psychological stress. Anxiety Stress Coping. 2019;32(6):711–27. 10.1080/10615806.2019.1648794. [DOI] [PubMed] [Google Scholar]
- 42.McEwen BS. Stress, adaptation, and disease. Allostasis and allostatic load. Ann N Y Acad Sci. 1998;840:33–44. 10.1111/j.1749-6632.1998.tb09546.x. [DOI] [PubMed] [Google Scholar]
- 43.Brosschot JF, Pieper S, Thayer JF. Expanding stress theory: prolonged activation and perseverative cognition. Psychoneuroendocrinology. 2005;30(10):1043–9. 10.1016/j.psyneuen.2005.04.008. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.