INTRODUCTION
Occupational burnout results from chronic workplace stress and is characterized by exhaustion, mental distancing from work, and reduced professional efficacy, with a high prevalence of 50% among physicians [1]. Beyond its mental health consequences, burnout has been prospectively associated with increased coronary heart disease (CHD) risk, independent of depression and traditional cardiovascular risk factors [2]. Poor sleep quality, which frequently co-occurs with burnout [3], is itself a well-established predictor of CHD [4]. Physicians are prone to sleep disturbance due to long hours, night shifts, and limited recovery, leading to fragmented and insufficient sleep [3].
Arterial stiffness, assessed by aortic pulse wave velocity (PWVao), is a robust marker of vascular aging and an independent predictor of CHD [5]. Poor sleep quality, measured by the Pittsburgh Sleep Quality Index (PSQI), has been associated with increased PWVao in older adults, independent of cardiovascular risk factors and mood symptoms [6]. Burnout has similarly been linked to increased arterial stiffness in young women [7], but evidence remains scarce, and potential interactive effects of burnout and sleep disturbance on arterial stiffness are unknown.
To address this gap and inform early prevention strategies, we examined whether burnout and poor subjective sleep quality, independently and interactively, are associated with PWVao in male physicians. The focus on men was deliberate, given their higher CHD rates compared to women [8] and the stronger association between occupational stress and CHD risk in men [9]. This study builds on our recent report in the same cohort, where poor sleep predicted impaired coronary microvascular function only in physicians with burnout [10]. Here, we test whether this pattern extends to arterial stiffness, a marker of large-artery health.
METHODS
Study Participants
As described previously [10,11], 60 practicing male physicians were enrolled through various channels for a larger observational study on cardiovascular health in physician burnout. Inclusion in the burnout group required an emotional exhaustion score ≥27 and/or depersonalization score ≥10 (with a minimum emotional exhaustion score of 20) on the 22-item Maslach Burnout Inventory-Human Services Survey. Controls had to score below 16 on emotional exhaustion and below 7 on depersonalization [1]. Additional criteria included moderate or lower depressive symptoms and at least 6 months of job-related stress in the burnout group, and no more than mild depressive symptoms in controls. This approach limited confounding by major depression while preserving real-world overlap between burnout and subthreshold depressive symptoms. All participants were non-smokers, aged 28–65 years, without known cardiovascular disease, diabetes, or major lifetime psychiatric disorders. Written informed consent was obtained from all participants, and the study was approved by the local ethics committee Zurich (BASEC No. 2018-01974).
Arterial Stiffness
PWVao (m/s) was measured noninvasively using the Arteriograph device (TensioMed Ltd.), a validated oscillometric system [12]. Measurements were taken in the sitting position after a 5-minute rest. The Arteriograph determines PWVao based on the time difference between the first (direct) and reflected pressure waves. Only recordings with quality criteria in the green range (standard deviation of PWVao <1.0 m/s) were included. For participants with two valid measurements, the mean value was used. PWVao is normal below 9.0 m/s.
Sleep Quality
Sleep quality over the past four weeks was assessed using the PSQI, a validated 19-item self-report instrument comprising seven components: subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction [13]. Each component is scored from 0 to 3, and the global score (range 0–21) reflects overall sleep quality, with higher scores indicating worse sleep. A cutoff score >5 distinguishes poor from good sleepers.
Covariates
Covariates, selected a priori for their known associations with arterial stiffness and/or sleep quality, included age (self-reported in years), body mass index (BMI, kg/m2; calculated from measured height and weight), self-reported habitual physical activity (times/week with sweat-inducing exercise), peripheral heart rate, and peripheral systolic blood pressure. Heart rate and blood pressure were measured twice with the Arteriograph device, and the averages were used in analyses.
Data Analysis
Data were analyzed using IBM SPSS Statistics 29.0 (IBM Corp.), with significance set at two-tailed p<0.05. Missing data (3.1% for PSQI, 3.3% for arteriography, and one poor-quality PWVao recording) were imputed using the expectation–maximization algorithm, as Little’s missing completely at random (MCAR) test (p=0.82) confirmed the MCAR assumption. Group differences were tested with the Mann–Whitney U test, and associations between continuous variables with Pearson correlations. Separate multiple linear regression models were run with normally distributed PWVao as the dependent variable. Predictors were mean-centered to reduce collinearity and improve interpretability, making the intercept reflect expected PWVao at average covariate values. Fully adjusted models included seven to eight theory-based predictors, a ratio that can raise overfitting risk given the sample size. Safeguards included collinearity checks, multivariate outlier screening (Mahalanobis distance, p<0.001), and interpretation based on statistical significance and effect size (partial r2). Model 1 examined burnout status as the sole predictor of PWVao, whereas Model 2 examined the PSQI global score as the sole predictor. Model 3 included both burnout status and the PSQI global score in the same model to test their independent, additive associations with PWVao. Model 4 extended Model 3 by adding the burnout×PSQI interaction term to test whether the association between sleep quality and PWVao differed by burnout status (moderation analysis). This sequential approach allowed for disentangling the main effects of burnout and sleep quality from their interactive effect.
RESULTS
Participant Characteristics
Table 1 summarizes participant characteristics by burnout status and sleep quality. One in four participants was classified as poor sleepers (PSQI global score >5), with a higher prevalence in the burnout group (40%) than controls (10%). Accordingly, the burnout group showed a higher mean PSQI global score and reported poorer subjective sleep quality, longer sleep latency, and greater daytime dysfunction. No significant group differences were found in BMI, physical activity, or cardiovascular measures, including PWVao. Poor sleepers did not differ from good sleepers in any characteristics other than PSQI component scores. In the full sample, global PSQI score was significantly associated with BMI (r=0.293, p=0.023), but not with any other clinical variable, including PWVao (p=0.94).
Group-Dependent Associations between Sleep Quality and Pulse Wave Velocity
Table 2 shows that neither burnout nor global sleep quality was independently associated with PWVao, whether examined separately (Models 1 and 2) or jointly (Model 3). In the fully adjusted Model 4, however, a significant burnout×PSQI interaction emerged (B=0.366; 95% confidence interval [CI], 0.095 to 0.636; p=0.009), accounting for 12.6% of the unique variance and suggesting clinical relevance. In other words, for each 1-point increase in PSQI score (poorer sleep), PWVao increased by 0.366 m/s more in burnout than in controls, consistent with Fig. 1A: the association was positive in burnout (B=0.197; 95% CI, −0.023 to 0.418; p=0.076) and negative in controls (B=−0.181; 95% CI, −0.376 to 0.014; p=0.067). Although neither within-group slope was significant, the interaction indicates a significant difference in slope between groups, with poorer sleep linked to steeper PWVao increases in burnout.
This group-dependent effect was driven by the PSQI sleep duration component, which showed a significant group×sleep duration interaction (B=0.926; 95% CI, 0.159 to 1.693; p=0.019; unique variance explained=10.3%). A 1-point increase in the PSQI sleep duration score (shorter habitual sleep) was associated with a 0.926 m/s greater PWVao in the burnout group than in controls. As shown in Fig. 1B, the associations were in opposite directions: positive in the burnout group (B=0.252; 95% CI, −0.383 to 0.887; p=0.42) and negative in controls (B=−0.671; 95% CI, −1.182 to −0.160; p=0.012), indicating that shorter sleep was linked to higher PWVao only in the burnout group. To better contextualize sleep duration, we extracted the self-reported values from participant files: 30 participants (50%) slept >7 hours/night in the past 4 weeks, 26 (43.3%) slept 6–7 hours, and 4 (6.7%) slept 5–6 hours.
No significant interactions between burnout status and any of the other PSQI components were observed in relation to PWVao. As expected, higher heart rate and older age were significantly associated with higher PWVao, whereas systolic blood pressure, BMI, and physical activity showed no significant associations with PWVao.
DISCUSSION
This study found that poorer, particularly shorter, sleep was associated with greater arterial stiffness, but only in physicians with occupational burnout. Importantly, the significant interaction between burnout and sleep quality means the difference in slopes between groups is statistically reliable, even though neither within-group slope alone reached significance; this is a common pattern when the relationship between two variables changes direction across subgroups. This suggests that psychosocial stressors may amplify the cardiovascular effects of sleep disturbances, a relationship previously highlighted as needing further study [14]. Our findings show that the impact on arterial stiffness is context-dependent, emerging only under high occupational stress, and, to our knowledge, are the first to report such an interaction in physicians. These findings align with our earlier report [10], linking poor sleep to impaired coronary microvascular function only in burnout, reinforcing that high occupational stress amplifies cardiovascular risk of poor sleep across vascular beds.
Short sleep duration emerged as the main driver of arterial stiffness. Sleep restriction can impair endothelial function, elevate sympathetic activity, and increase arterial stiffness [14]. Our findings extend this evidence by showing that burnout may magnify these effects even in a relatively healthy, professionally active cohort. The effect was clinically meaningful: in our sample, reducing sleep from >7 hours to 6–7 hours/night corresponded to a 0.93 m/s higher PWVao in burnout compared to controls. A 1 m/s PWVao increase has been linked to 14%–15% higher risk of cardiovascular events and mortality, likely through increased left ventricular afterload, hypertrophy, and impaired coronary perfusion [15]. Similarly, a 3-point higher PSQI global score corresponded to a 1.1 m/s higher PWVao in burnout compared to controls.
Neither burnout nor sleep quality alone was associated with PWVao in the full sample, likely reflecting careful matching on cardiovascular risk factors, the relatively healthy and functionally preserved status of participants, and low variability in cardiovascular risk. The significant burnout-by-sleep quality interaction indicates that poor sleep predicted greater stiffness only in the presence of burnout, suggesting that high psychosocial stress amplifies vascular impact of sleep disturbance even in the absence of overt disease. Interestingly, in controls, shorter sleep duration was significantly associated with lower PWVao, and poorer global sleep quality showed a similar trend. Although speculative, these inverse patterns may reflect physiological resilience, chance variation given the modest sample size, or sample-specific effects and warrant replication in larger studies.
Clinically, these results support integrated cardiovascular risk management in high-stress professions. Routine sleep assessment in individuals with burnout may help identify those at increased vascular risk. Interventions targeting sleep hygiene, work-hour regulation, and stress reduction could help mitigate long-term cardiovascular burden. Relevant to our findings, weekend catch-up sleep has been linked to lower burnout risk in medical staff with weekday sleep <7 hours [16]. Objective vascular measures such as PWVao could further refine occupational health screening when burnout and sleep complaints co-occur.
Limitations include the cross-sectional design, modest sample size, male-only sample, and reliance on subjective sleep measures. Nevertheless, the findings highlight that poor sleep quality, especially short sleep duration, was associated with increased arterial stiffness in physicians with burnout. The findings emphasize the importance of assessing and improving sleep health in high-stress occupational settings, both for psychological well-being and for cardiovascular prevention. Future studies should aim to replicate these findings in larger, mixed-gender cohorts, ideally incorporating objective sleep metrics.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
