Participants with and without coronary artery disease (CAD), hypertension, or atrial fibrillation were recruited from three tertiary hospitals in metropolitan Adelaide, South Australia (The Queen Elizabeth Hospital, the Royal Adelaide Hospital, and the Lyell McEwin Hospital), medical centers based in Sydney, New South Wales, and the general Adelaide population. Eligibility included age 18 years or older, English language proficiency, and willingness to complete the 24-hour blood pressure monitoring. Participants were ineligible if they had a history of a recent transient ischemic attack (TIA), stroke, acute myocardial infarction (MI), or decompensated heart failure in the past six months, dementia or neurodegenerative conditions (e.g. Parkinson’s Disease), traumatic brain injury, or significant head trauma, a diagnosis of any life-limiting medical condition likely to be fatal within a year, alcohol or substance use disorders, psychosis or other severe psychiatric conditions requiring hospitalization, or a high risk of self-harm. The aim was to include a heterogeneous sample reflective of real-world cardiovascular risk profiles across clinical and non-clinical populations obtained from cardiology out-patient, primary care, and general populations.

Six of the 94 participants were excluded for having severe psychiatric conditions (n = 2) or being unable to wear the ABPM for 24 hours (n = 4). The final sample size was 88. Participation was voluntary, and those recruited from the general community were compensated with a Visa gift card of $50.00. This study was approved by the Human Research Ethics Committees of Central Adelaide Local Health Network (CALHN No 11733), The University of Adelaide (HREC/15/TQEH4), and the University of New England (Approval No HRE23-004).

All participants received an information sheet and completed eligibility screening via phone or electronically. Written consent and demographic information were obtained, followed by an eligibility screener, the DSM-5 Self-Rated Level 1 Cross-Cutting Symptom Measure – Adult (DSM-5 XC), used to screen for alcohol or substance use disorders, psychosis, or other psychiatric conditions in individuals as part of the exclusion criteria process. The 23-item questionnaire assessed 13 psychiatric domains, including depression, anger, mania, anxiety, somatic symptoms, suicidal ideation, psychosis, sleep problems, memory, repetitive thoughts and behaviors, dissociation, personality functioning, and substance use (32). The DSM-5 XC is clinically validated in screening for psychiatric disorders and has good test-retest reliability in the DSM-5 Field Trials conducted in adult clinical samples internationally (32). Participants were excluded if they scored ≥ 3 for modules of self-harm, psychosis, or alcohol and drug abuse and were referred to their general practitioner as part of duty of care. Eligible participants then proceeded to complete measures of cognitive function (not reported further), depression, anxiety, and blood pressure (See Figure 1 ).

The Generalized Anxiety Disorder scale (GAD-7) (33), is a 7-item brief questionnaire that be used to measure symptoms of anxiety. Items are based on the diagnostic criteria for Generalized Anxiety Disorder defined in the DSM-5 (32). This includes symptoms of nervousness, uncontrollable and excessive worry, difficulty relaxing, restlessness, irritability, and feeling afraid. Each item in the GAD-7 is scaled from “0” (not at all) to “3” (nearly every day), with sum scores representing severity. The GAD-7 scale can be effectively divided into cognitive-emotional and somatic components. Cognitive questions focus on worry and mental tension, while somatic questions address physical symptoms like restlessness and irritability (34). The scale has been found to have good internal consistency (α = 0.89) (33).

As per the screening recommendations of the American Heart Association (35), the Patient Health Questionnaire (PHQ-9) (36), was used to quantify depressive symptoms. The PHQ-9 is a 9-item self-administered scale that measures symptoms requisite for the diagnosis of depression per the DSM-5 (32). Symptoms measured include diminished interest or pleasure, hopelessness, sleep difficulties and fatigue, changes in appetite or weight, feeling like a failure, concentration difficulties, psychomotor agitation, and suicidal ideations. Like the GAD-7, each item is scaled from “0” (not at all) to “3” (nearly every day), with sum scores representing severity. Similar to the GAD-7, the PHQ-9 has a bi-dimensional structure, with somatic (sleep problems, fatigue, appetite problems, and psychomotor difficulties) and cognitive-affective (lack of interest, depressed mood, negative feelings about self and concentration) dimensions (37). The PHQ-9 is a short and simple measure with good psychometric properties, including construct and criterion validity and internal consistency (α of 0.89) (36). Both the GAD-7 and PHQ-9 have been used in Australian cardiovascular populations including cardiac surgery (38), heart failure (39), and general CVD populations.

The Sleep Condition Indicator (SCI) is an eight-item self-report scale developed to screen for insomnia disorder in line with DSM-5 criteria (40) as well as sleep quality and insomnia symptoms in research and clinical settings. This instrument assesses key areas related to sleep difficulties, including problems with initiating and maintaining sleep, subjective sleep quality, impact on daytime functioning and performance, duration and frequency of sleep disturbances, and distress caused by poor sleep. Each item is rated on a five-point scale from 0 to 4 (40). The SCI demonstrates excellent psychometric properties, with robust internal consistency (α ≥ 0.86) and convergent validity with the Pittsburgh Sleep Quality Index and the Insomnia Severity Index. The SCI was reverse scored to align with the directionality of the GAD-7 and PHQ-9 so that higher SCI scores indicated greater severity of sleep disturbance and poorer functioning (41).

Statistical analyses were performed using IBM SPSS Statistics (Version 29; IBM Corp, 2023). Data normality was assessed using the Shapiro-Wilk test. Extreme BP readings were screened, with exclusion thresholds set at systolic >240 mmHg or <50 mmHg and diastolic >140 mmHg or <40 mmHg; no data met these exclusion criteria. Spearman’s correlations were conducted to assess the relationships between central and brachial BPV and individual symptom items and total scores from the GAD-7, PHQ-9 and SCI questionnaires across ambulatory, awake, and asleep periods. Multiple linear regressions were employed to examine the independent relationships between BPV metrics and these selected factors across all three time periods. Model 1 assessed unadjusted associations, while Model 2 accounted for potential confounders by adjusting for age and gender. Outliers were identified using standard residuals (≥ 3); however, none were detected. Subsequently, mediation analyses were conducted to investigate the mediating role of sleep in the relationship between BPV and mood outcomes. Given the cross-sectional nature of the data, mediation models were used for exploratory purposes to examine potential indirect associations, without inferring causal or temporal direction. Using the PROCESS macro of SPSS, bootstrapping with 10,000 resamples was applied to estimate confidence intervals for indirect effects. The regression model (unadjusted or adjusted) with the most consistent associations was chosen for the mediation analysis, ensuring a robust exploration of BPV and mood outcomes. As an exploratory study, no corrections for multiple comparisons were applied (42, 43).

The final sample comprised 88 participants (females n = 34, 39%) across South Australia (n = 69) and Sydney (n = 19). The median age was 57.6 years old (range 23 – 86). Participants’ body mass index (BMI) ranged between healthy and obese. More than half of the participants were employed or self-employed (56%). Approximately 67% of participants were non-smokers, and most had a university degree. The most prevalent cardiovascular disease among the population was hypertension (32%), followed by coronary artery disease (21%) and atrial fibrillation (11%) (see Table 1 ), supporting generalizability to mid-aged adults with and without cardiovascular risk.

Figure 5 illustrates the mediating role of sleep quality (sleep patterns and sleep-related issues) in the BPV-GAD-7 symptom relationship, with adjustments for age and gender. Sleep-related issues emerged as a stronger mediator for both somatic and cognitive GAD-7 symptoms. Among central BPV metrics, AP and AIx were linked to better sleep-related functioning, which mediated their association with lower anxiety symptoms. AIx showed significant indirect effects on somatic (β = -0.01, p < 0.05) and cognitive (β = -0.02, p < 0.05) symptoms, while AP exhibited similar effects, particularly for cognitive symptoms (β = -0.02, p < 0.05). Conversely, brachial BPV metrics (MAP, diastolic BPV) were linked to worse sleep-related functioning, with brachial MAP showing significant indirect effects on cognitive symptoms (β = 0.06, p < 0.05), but no mediation effects for somatic symptoms. Sleep pattern was significantly associated with central PP for both GAD-7 symptom types, with higher PP linked to worse sleep patterns (Tables 5, 6). However, no significant indirect effects were observed, suggesting sleep patterns play a weaker mediating role compared to sleep-related functioning issues.

This study supports evidence that central BPV is associated with both cognitive and somatic anxiety symptoms, suggesting a common underlying mechanism involving autonomic and cerebrovascular dysfunction. While ANS disruptions are traditionally linked to somatic symptoms such as breathlessness and restlessness (11) and CBF instability to cognitive symptoms like excessive worry and impaired concentration (12), the similar BPV associations across symptom dimensions challenge this distinction. Instead, these findings suggest that BPV-related autonomic and cerebrovascular dysregulation may contribute broadly to anxiety symptoms rather than selectively to somatic or cognitive domains.

However, these findings contrast with most prior studies, which have reported positive associations between BPV and anxiety and have attributed this relationship to autonomic dysregulation, increased sympathetic activity, and reduced heart rate variability (31, 44–47). One potential explanation for these discrepancies is the variation in BPV measurement methods across studies. Many previous investigations have focused on brachial BPV, which may not fully capture cerebrovascular and autonomic contributions to anxiety. The current study’s inclusion of both central and brachial BPV offers a more comprehensive perspective due to its proximity to vital organs (22), reinforcing the importance of central BPV as a more physiologically relevant marker (20, 21). Differences in BPV measurement timeframes may also contribute to inconsistencies. Zhou et al. (2021) examined visit-to-visit BPV over a three-year period (47), whereas Virtanen et al. (2003) measured beat-to-beat BPV over just five minutes (44), neither of which accounted for nocturnal BPV or fluctuations over a 24-hour ambulatory period. The absence of circadian BP regulation in some studies may overlook key autonomic and cerebrovascular changes that influence anxiety symptoms (18, 25).

Longitudinal research further complicates the BPV-anxiety relationship. While acute anxiety is associated with increased BPV, likely due to baroreflex impairment, chronic anxiety has been linked to lower systolic BP over time, potentially leading to attenuated BPV (44, 48). This distinction underscores the importance of differentiating between acute and chronic anxiety in BPV research. Overall, these findings suggest that BPV-anxiety relationships depend on BPV type and symptom specificity. While brachial BPV and global anxiety scores have been linked to higher BPV, our results suggest that central BPV’s AIx, may reflect distinct autonomic patterns related to anxiety symptomatology. The difference in findings between central and brachial BPV may reflect underlying physiological differences. Central BPV, influenced by large artery stiffness and wave reflections, is more directly tied to cerebral perfusion and baroreflex sensitivity; key mechanisms implicated in emotional regulation and anxiety (44). In contrast, brachial BPV reflects peripheral pressure fluctuations that may be dampened or distorted due to arterial branching and muscular resistance (20). These differences underscore the importance of considering arterial measurement site when interpreting BPV-related associations, particularly for symptoms modulated by central autonomic networks.

Unlike anxiety, BPV-depression associations were inconsistent, with no significant links to cognitive or somatic symptom dimensions. This finding does not align with research suggesting that BPV influences mood through autonomic and cerebrovascular mechanisms, particularly given the ANS’s role in physiological regulation and CBF’s involvement in cognitive and emotional processing (11, 12). However, when examining individual symptoms, hopelessness and psychomotor agitation were negatively associated with central BPV. This aligns with prior research indicating that hopelessness, independent of total depressive scores, is associated with BPV and vascular dysfunction (29). These findings suggest that examining individual depressive symptoms may provide greater insight into BPV-depression associations than broader cognitive and somatic dimensions.

Prior research remains inconsistent on the association between BPV with depression symptoms. Some studies propose that higher BPV contributes to depressive symptoms via cerebral microvascular dysfunction, autonomic instability, and cerebrovascular dysregulation (49, 50). However, others found no significant BPV differences between depressed and non-depressed individuals, suggesting a weaker or context-dependent BPV-depression link (44, 51). Differences in BPV measurement methods may partly explain the lack of consistent findings in previous research. Long-term BPV, particularly visit-to-visit BPV, reflects chronic vascular instability and has been more strongly linked to depression than short-term (24-hour) BPV (50). This suggests that prolonged exposure to high BPV, rather than short-term fluctuations, may contribute to depressive symptoms through mechanisms implicated in the vascular-depression hypothesis, which associates late-onset depression with cerebrovascular dysfunction and cognitive decline (50).

Demographic and lifestyle factors further complicate BPV-depression associations. Huang et al. (2020) found that higher systolic BPV was linked to depressive symptoms only after adjusting for BMI, smoking, and alcohol use (52), while Vallee et al. (2020) reported that lower BP and hypertension correlated with depression, with education level acting as a key modifying factor in women (53). These findings emphasize the need to control for individual differences when examining BPV in depression. While this study found limited evidence linking BPV to depression, the observed item-level associations suggest that future research should prioritize individual symptom analysis and long-term BPV measures comprehensively understand BPV’s role in depressive symptomatology.

Our findings indicate that BPV is associated with sleep disturbances, particularly sleep-related functioning rather than overall sleep quality. Lower central BPV metrics (SBP, MAP, and AIx) were linked to worse sleep-related functioning, while higher PP was associated with greater sleep disturbances, including frequent awakenings and poor sleep efficiency. These findings highlight distinct BPV metrics influencing different aspects of sleep dysfunction, reinforcing the need for granular BPV analysis. The association between higher PP and greater sleep disturbances aligns with evidence that arterial stiffness and wave reflection contribute to autonomic instability during sleep (54, 55). Similarly, sleep apnea, characterized by nocturnal BP oscillations, has been shown to amplify BPV, linking disrupted sleep architecture to heightened cardiovascular reactivity (56). Conversely, the relationship between lower BPV and increased sleep dysfunction may reflect reduced physiological flexibility in nocturnal BP regulation, as seen in individuals with insomnia and sleep apnea (57, 58).

The partial mediation of BPV-anxiety associations by sleep-related issues suggests that poor sleep functioning amplifies BPV-related anxiety symptoms, consistent with findings that sleep disturbances heighten sympathetic activation and cardiovascular reactivity (59). Mediation effects observed for AP and AIx indicate that arterial wave reflection may be a key hemodynamic mechanism linking sleep dysfunction to anxiety.

These findings align with evidence that poor sleep quality contributes to increased BPV (13) and that mood states influence nocturnal BP regulation, with negative moods (e.g., anger, sadness) linked to higher night time BP (60). Additionally, irregular sleep patterns have been associated with increased BP and hypertension risk, reinforcing the role of consistent sleep in cardiovascular stability (61). The absence of a mediation effect for depression supports the hypothesis that BPV is more relevant to autonomic regulation in anxiety than depression. Future research should investigate whether altered nighttime BP patterns contribute to long-term anxiety symptom severity and cardiovascular risk. These findings also have relevance for public health and clinical care. Central BPV may serve as a non-invasive marker of psychological distress in cardiovascular populations and inform early screening strategies. Moreover, interventions targeting sleep and autonomic regulation—such as Cognitive Behavioural Therapy for Insomnia (CBT-I) physical activity, and stress reduction—may help reduce BPV and anxiety symptoms, improving outcomes on both physiological and psychological fronts (54).

This study presents several methodological strengths including the use of central BPV, which may provide a more accurate representation of cerebrovascular regulation and autonomic function compared to brachial BPV (20, 23). By incorporating multiple BPV indices over a 24-hour period, including MAP, PP and AIx, this study offers a comprehensive evaluation of short-term BP fluctuations in relation to depression, anxiety and sleep, and any possible physiological processes (62). Furthermore, 24-hour ABPM was utilized, which is superior to clinic-based or single-timepoint BP assessments, as it reduces white coat effects and captures real-world BP fluctuations that are more reflective of autonomic regulation (63). Another strength of this study is its symptom-specific analysis, differentiating between somatic and cognitive symptoms. Prior research has often generalized mood disturbances, which may obscure symptom-specific associations with BPV (29, 64).

A key limitation of this study is that analyses were largely exploratory in nature, and therefore no corrections for multiple comparisons were applied. While this may increase the risk of false positives, sensitivity in detecting potential associations was prioritised to inform hypotheses for future research (42, 43). Analyses were also constrained to using the CV, which does not capture the temporal association between BP measurements. In contrast, the average real variability (ARV) method takes into consideration the interrelated nature of successive measurements and is also associated with cognitive decline and cardiovascular risk (63, 65). Another limitation when using a 24-hour ambulatory BP monitoring to measure nocturnal BPV in a study where sleep was also quantified, is that the monitor itself may contribute to disrupted sleep and poorer sleep quality whilst simultaneously contributing to variance in nocturnal BP readings. Frequent disturbances from BP monitoring can lead to artificial fluctuations in nighttime BP measures, potentially influencing sleep-dependent autonomic regulation (57).

This study also did not standardize physical activity or dietary intake, both of which influence BPV, mood, and sleep. Caffeine intake, in particular, has been shown to increase BP fluctuations, potentially affecting the cardiovascular system’s regulatory mechanisms (66). Additionally, consuming 200–300 mg of caffeine has been shown to increase systolic BP by approximately 8.1 mmHg and diastolic BP by 5.7 mmHg, with effects lasting at least three hours (67, 68). The lack of control over dietary intake, including caffeine consumption, may have introduced unaccounted variability in BPV findings. Similarly, while hypertension medication use was recorded, it was not statistically adjusted for. Antihypertensive medications affect vascular tone, autonomic responses, and BPV, potentially influencing BPV-mood relationships (69). Future research should incorporate a more comprehensive set of covariates to account for confounding factors.

Additionally, the study was constrained by a small sample size, reducing statistical power and increasing the risk of Type II errors (70). The broad age range (23–86 years, median 57.6 years) may also dilute age-specific effects, particularly as BPV, mood, and cognitive function are known to vary across the lifespan (47). While age was included as a covariate in adjusted models, age-stratified analyses may provide greater specificity in detecting BPV-mood interactions. Gender differences were also not explicitly analyzed, and it has been shown that BPV-mood relationships may differ between males and females due to hormonal and vascular factors (47). Vallee et al. (2020) reported that lower BP and hypertension correlated with depression, with education level acting as a key modifying factor in women (53, 71). These findings emphasize the importance of controlling for demographic and lifestyle factors when examining BPV-psychological health interactions. These limitations may affect the strength and generalizability of our conclusions. Future studies should use longitudinal designs to clarify temporal relationships, apply age- and gender-stratified analyses, and control for lifestyle factors such as caffeine, physical activity, and medication adherence to reduce confounding and enhance causal interpretation (63, 69).

Finally, the cross-sectional design significantly limits causal inferences. This is particularly relevant to the mediation analyses, which assume a temporal sequence between variables that cannot be confirmed within a cross-sectional framework. Longitudinal research is needed to clarify whether BPV changes precede or follow mood disturbances and to establish bidirectional relationships between BPV, sleep, and psychological health (72, 73). Future studies should also integrate alternative BPV metrics (e.g., ARV and HRV) to enhance measurement precision. Interventional research targeting BPV modulation—via exercise, dietary changes, and stress reduction techniques—could further clarify the association between BPV and psychological distress. Expanding statistical models through propensity score matching and structural equation modelling will improve control for unmeasured confounders (e.g. education level, BMI, caffeine intake and antihypertensive use). Additionally, age- and gender-stratified analyses and recruitment from diverse community-based populations will improve generalizability. By addressing these gaps, future research can refine BPV’s role as a dynamic marker of psychological and autonomic regulation and explore its potential as a therapeutic target for mood and sleep disturbances.

This study indicates that central BPV showed stronger and more consistent associations with anxiety symptoms than depression, reinforcing its relevance in autonomic and cerebrovascular dysregulation. In contrast, BPV-depression associations were weak and inconsistent, suggesting that BPV may be more relevant to individual depressive symptoms, such as hopelessness and psychomotor agitation, rather than broader symptom dimensions. Additionally, sleep disturbances, particularly sleep-related functioning, mediated the BPV-anxiety relationship, highlighting the role of sleep in autonomic regulation and psychological distress. By incorporating central BPV metrics and circadian BPV patterns, this study advances understanding of the physiological mechanisms linking BPV to mental health. Future research should employ longitudinal designs to clarify causal pathways, explore alternative BPV metrics such as visit-to-visit variability, and examine sex- and age-specific effects. Investigating additional confounders, including medication and lifestyle factors, may further refine BPV’s role in psychological distress. Given the mediating role of sleep, future research should investigate whether interventions targeting sleep quality and autonomic regulation could benefit both mental and cardiovascular health.