BPV refers to the magnitude or degree of fluctuations in blood pressure over a period of time. It describes the changes in blood pressure values around the mean, rather than a fixed blood pressure value. BPV can be assessed over different time frames: very short-term (sequential measurements), short-term (through ambulatory blood pressure monitoring over 24 h), medium-term (daily home blood pressure monitoring), or long-term (measured at each clinical visit, which can span weeks to years) (Table 1).

2.2.1 Dispersion[Standard Deviation (SD), Coefficient of Variation (CV), Variability Independent of Mean (VIM)]. SD is the most commonly used metric for measuring BPV, related to the mean blood pressure value and influenced by environmental stressors and circadian variations. Weighted SD is used exclusively for short-term BPV measurement, calculating SD for daytime and nighttime separately and averaging them with weights to eliminate the impact of nocturnal blood pressure decline on SD. CV is used to adjust the relationship between SD and mean blood pressure. CV has stronger prognostic capability than SD because it can identify individuals whose BPV exceeds expected ranges (1). VIM is a transformation of SD that eliminates the influence of mean blood pressure on BPV through nonlinear regression; however, its parameters are specific to each cohort and cannot be compared across populations.

2.2.2 Sequence[Average Real Variability (ARV)]. ARV is not influenced by circadian variations or mean blood pressure and is defined as the average of the absolute differences between consecutive blood pressure measurements. It is considered more suitable for measuring 24-hour BPV, as ARV can reveal trends in blood pressure fluctuations during continuous measurements. Compared to SD, it is a more reliable prognostic indicator, as it is more sensitive to individual blood pressure measurement sequences and less sensitive to low sampling frequencies (1).

2.2.3 Instability[Range (Maximum—Minimum BP), Peak Size (Maximum BP), and Trough Value (Mean—Minimum BP)]. The range is the difference between the maximum and minimum values, used to quantify short-term, medium-term, and long-term BPV, and is easily influenced by extreme blood pressure values.

2.2.4 Frequency[Residual Standard Deviation (RSD)]. If there is a linear relationship between time and blood pressure, BPV can be defined as RSD, which is the root mean square error of the difference between predicted and observed blood pressure. This measure is less affected by changes in blood pressure over time, but the assumption of a linear trend does not apply in all situations.

Weighted SD, CV, and ARV were employed to overcome the limitations of conventional SD measurements, providing more accurate evaluations and improved prediction of target organ damage (1). Despite the diversity of assessment methodologies, there remains a notable absence of a unified “gold standard” for BPV quantification, which has contributed to substantial heterogeneity across studies. Consequently, it is imperative to accurately and consistently identify the most reliable metrics while meticulously documenting blood pressure measurements in research protocols. In clinical practice, the selection of appropriate parameters should be guided by individual patient characteristics, with ARV demonstrating particular utility for elderly patients or those with cardiovascular comorbidities.

Intraoperatively, arterial catheters (radial artery cannulation) are used to achieve real-time arterial blood pressure monitoring. Ambulatory blood pressure monitoring (ABPM) is employed to assess short-term blood pressure fluctuations over a 24-hour period, allowing for unrestricted monitoring of changes at numerous time points during surgery (1). This method can provide high-frequency blood pressure data, typically obtaining a reading every 5 s. Non-invasive cuff measurements can also be used; however, although this method is more convenient, its relatively low measurement frequency cannot fully capture dynamic blood pressure changes, and many brief episodes of high and low blood pressure may be missed with intermittent measurement. The blood pressure monitor records blood pressure data throughout the surgical procedure, which are usually displayed and saved in the form of waveforms. These data can then be utilized for subsequent statistical analysis and evaluation.

In normal physiological conditions, BPV is primarily influenced by factors such as ventilation, sympathetic nervous system activity, circadian fluctuations in cortisol secretion, increased vascular sensitivity to norepinephrine, diurnal patterns of renin and aldosterone activity, as well as physical activity and sleep habits (4). Under pathological conditions, autonomic dysfunction, persistent activation of the Renin–Angiotensin–Aldosterone System (RAAS), arterial stiffness, and endothelial dysfunction may impair the buffering capacity of the vascular bed against pressure fluctuations, thereby establishing a vicious cycle in which elevated BPV promotes target organ damage and vice versa (31).

The determinants of ARV are multifactorial, involving baseline patient characteristics, perioperative stress, and anesthesia management. Age is one such important factor; ARV tends to increase with aging, suggesting that elderly patients may experience greater hemodynamic variability during anesthetic recovery. In addition, intraoperative blood loss is a key contributor, as greater blood loss predisposes patients to instability during recovery. Factors such as duration of surgery, degree of surgical stimulation, changes in intravascular volume, and certain anesthetic drugs or techniques may also exert direct or indirect influences on ARV (5). Previous studies have indicated that the use of dexmedetomidine, spinal anesthesia or combined regional blocks, and relatively higher doses of opioids in selected high-risk patients can attenuate sympathetic excitation triggered by surgical stimuli, thereby reducing intraoperative blood pressure and heart rate fluctuations and limiting excessive increases in BPV (33–35). Anesthetic depth likewise contributes to IBPV; excessive anesthesia may suppress autonomic responses and generally leads to sympathetic inhibition (6). Evidence further shows that ARV is significantly higher in certain populations, including female patients, individuals with diabetes, smokers, and those with peripheral vascular disease, atrial fibrillation, or a history of transient ischemic attack (TIA) or stroke (7). Among these, patients with autonomic neuropathy (e.g., diabetes or Guillain–Barré syndrome) are particularly vulnerable due to impaired sympathetic–parasympathetic balance and baroreceptor reflex dysfunction. Their diminished ability to adjust to hemodynamic stressors such as anesthetic agents, volume shifts, and positional changes places them at increased risk of blood pressure fluctuations and elevated ARV during the perioperative period (30). These factors, closely linked to patients' physiological and pathological characteristics, contribute to greater uncertainty and heightened risk during anesthetic recovery.

In recent years, IBPV has emerged as an important perioperative determinant of postoperative delirium (POD). Unlike absolute hypotension or the duration of low mean arterial pressure (MAP), fluctuations in blood pressure more accurately reflect the integrated response of autonomic regulation and cerebral autoregulatory capacity under the stress of anesthesia and surgery. Consequently, IBPV serves as a more sensitive indicator of transient cerebral perfusion instability. A growing body of evidence across various surgical populations has consistently demonstrated a significant association between IBPV and the development of POD (Table 2).

In elderly patients undergoing hip fracture surgery, Zhang et al. conducted a retrospective cohort analysis involving 963 cases and found that an intraoperative mean arterial pressure variability (MAPV) exceeding 2.17% markedly increased the risk of POD (OR = 2.379, 95% CI: 1.496–3.771, P < 0.001). This association remained robust after propensity score matching (OR = 2.851, P < 0.001) (8). The authors proposed that excessive blood pressure fluctuations may induce repeated episodes of cerebral hypoperfusion, leading to microcirculatory dysfunction and metabolic derangements, thereby precipitating delirium more so than hypotension alone.

Comparable findings have been reported in cardiac surgical populations. Shen et al., in a cohort of 2,164 patients undergoing cardiopulmonary bypass, identified a bidirectional relationship between perioperative ARV and POD risk. Both a markedly low intraoperative ARV and an excessively high systolic ARV within 24 h postoperatively were independent predictors of delirium (OR = 1.17, P = 0.002) (9). These findings suggest that overly stable hemodynamics may signify impaired cerebrovascular reactivity, whereas excessive postoperative BPV can further disrupt autoregulatory integrity. Both extremes may culminate in cerebral hypoperfusion heterogeneity and neural dysfunction.

The deleterious effects of IBPV are particularly pronounced in patients with compromised cerebrovascular reserve. Liu and He demonstrated that elevated ARV-MAP was an independent risk factor for emergence delirium in pediatric patients with moyamoya disease (OR = 9.17, 95% CI: 3.85–21.82, P < 0.001), accompanied by a higher incidence of early postoperative ischemic events (26). This finding underscores that even brief fluctuations in blood pressure can provoke perfusion-metabolism mismatch and neuroinflammatory cascades in individuals with impaired autoregulatory capacity.

Collectively, evidence from cardiac, orthopedic, and neurovascular surgeries converges to highlight IBPV as a critical determinant of POD. The underlying mechanisms are multifactorial, involving disruption of cerebral autoregulation, blood-brain barrier instability, and microcirculatory oxygen imbalance. Excessive fluctuations may trigger ischemia-reperfusion injury and inflammatory activation, whereas excessive suppression of variability may mask inadequate perfusion responsiveness. Both conditions can ultimately impair neuronal integrity and cognition.

From a clinical standpoint, perioperative hemodynamic management should shift focus from merely “maintaining blood pressure levels” to “controlling blood pressure variability.” Efforts should be directed toward minimizing abrupt hemodynamic oscillations while ensuring adequate perfusion pressure. Integrating continuous cerebral oxygen monitoring, individualized fluid optimization, and tailored depth of anesthesia control may collectively mitigate the risk of POD and improve postoperative neurocognitive outcomes.

IBPV reflects the stability of hemodynamic regulation under surgical and anesthetic stress. Excessive fluctuations in blood pressure can lead to uneven renal microcirculatory perfusion and ischemia–reperfusion injury, thereby serving as a potential trigger for postoperative acute kidney injury (AKI). Unlike sustained hypotension, even when MAP remains within a normal range, frequent oscillations in blood pressure may disrupt renal autoregulation and increase the risk of renal impairment (17, 18).

A growing body of evidence from cardiac surgical populations has demonstrated a significant association between elevated BPV and postoperative renal dysfunction. Jinadasa et al. analyzed 3,687 patients undergoing CPB and reported that each 0.10 increment in the CV-SBP increased the odds of renal failure by approximately 104% (OR = 2.04, 95% CI 1.33–3.14), with the strongest impact observed during the prebypass period; in contrast, variability in MAP did not independently predict renal outcomes (20). Similarly, Packiasabapathy et al. employed Poincaré plots and coefficients of variation to quantify IBPV and found that, while these parameters captured dynamic hemodynamic changes, their predictive accuracy for AKI remained limited, suggesting that BPV should complement rather than replace established perioperative risk models (16). In liver transplantation, the influence of BPV appears even more pronounced. Xu et al. investigated 437 transplant recipients and identified cumulative duration and area of MAP < 65 mmHg, ARV, and VIM as independent predictors of post-transplant AKI (OR = 1.051, 1.008, 1.149, and 2.430, respectively; all P < 0.01), whereas greater intraoperative urine output and administration of terlipressin were protective factors (17). Pediatric patients, whose kidneys are developmentally immature and possess limited perfusion reserve, exhibit heightened vulnerability to BP fluctuations. Fishbein et al. reviewed 231 pediatric cardiac surgical cases and found that 51.5% developed postoperative AKI, with both systolic and diastolic ARV independently associated with its occurrence (OR ≈ 1.4, P < 0.05), particularly among infants younger than 12 months (18). More recently, Xiao et al. demonstrated in a cohort of children aged 0–7 years that MAP excursions exceeding ±30% of baseline during CPB, quantified as the cumulative area under the curve (±30%AUCm), were significantly correlated with AKI (P = 0.038), indicating that excessive pressure oscillations may surpass the threshold of renal autoregulation (19).

Collectively, these studies across diverse surgical contexts consistently highlight that greater IBPV is strongly associated with a higher incidence of postoperative AKI (Table 3). The underlying mechanisms likely involve impaired renal autoregulation, oxidative stress, and microcirculatory injury. Although the quantitative thresholds and metrics used to define BPV (e.g., SD, CV, ARV, VIM) vary across studies, maintaining hemodynamic stability throughout surgery has emerged as a key strategy for renal protection. Future investigations should integrate continuous hemodynamic monitoring with individualized anesthetic management to establish optimal BPV control targets and thereby mitigate the risk of postoperative AKI.

IBPV has emerged as an important hemodynamic marker associated with early postoperative outcomes, particularly short-term mortality. Large-scale studies in both cardiac and non-cardiac surgery have demonstrated that excessive intraoperative fluctuations in blood pressure, independent of MAP levels, significantly increase the risk of death within 30 days postoperatively.

Aronson et al. analyzed 7,504 patients undergoing coronary artery bypass grafting (CABG) and found that greater SBP excursions beyond the optimal range of 105–130 mmHg were strongly predictive of 30-day mortality (OR 1.03 per minute, 95% CI 1.02–1.39, P < 0.0001). Among various indices of BPV, the mean duration of systolic excursions outside this range exhibited the highest predictive power, suggesting that both the magnitude and persistence of BP deviations are crucial determinants of adverse outcomes (22). Subsequent studies expanded upon these findings by assessing BPV using more quantitative indices such as the CV. Jinadasa et al. conducted a retrospective study of 3,687 patients undergoing cardiac surgery with cardiopulmonary bypass and demonstrated that an increase of 0.10 in the SBP-CV was associated with a 150% increase in 30-day mortality (OR 2.50, 95% CI 1.60–3.92, P < 0.0001). Notably, this effect was driven primarily by the prebypass phase rather than the postbypass period, implying that hemodynamic instability early in surgery exerts disproportionate influence on postoperative survival (20). Similar associations have been observed across broader surgical populations. Benolken et al. analyzed 2,658 adult patients undergoing a variety of major non-cardiac procedures and reported that both preoperative and IBPV independently correlated with adverse outcomes, including 30-day mortality and major complications (adjusted OR 1.42–1.65, P < 0.01). Importantly, their study highlighted that BPV exerts prognostic influence regardless of baseline comorbidities or mean pressure levels, reinforcing its potential role as a physiologic marker of cardiovascular resilience (10). In contrast, variability in MAP (CV-MAP) did not independently predict mortality, indicating that SBP dynamics may be more sensitive to physiological stress during anesthesia induction and surgical manipulation. While these results underscore the detrimental effect of large BP oscillations, other investigations have yielded more nuanced conclusions. Packiasabapathy et al. analyzed a similar cohort using nonlinear Poincaré plots and BPV-derived indices (SD1, SD2, CV) and reported that these parameters alone did not significantly improve mortality prediction beyond established risk models such as the Society of Thoracic Surgeons score (C-statistic ≈ 0.7 for both). This suggests that although BPV reflects intraoperative instability, it may serve better as a complementary physiological marker rather than a standalone predictor of outcome (16).

Collectively, the current evidence indicates that elevated IBPV—especially sustained systolic deviations outside individualized optimal ranges—correlates strongly with higher 30-day mortality following cardiac and major non-cardiac surgery (Table 4). The predictive effect appears most pronounced during the early (prebypass) phase, highlighting the importance of early hemodynamic control. However, the integration of BPV metrics into perioperative risk assessment should be contextualized alongside traditional scoring systems and patient comorbidities. Future prospective studies are warranted to determine whether real-time BPV monitoring and targeted anesthetic interventions can effectively reduce perioperative mortality risk.

In recent years, accumulating evidence has underscored the prognostic importance of IBPV in neurosurgical procedures. This issue is particularly salient in revascularization surgery for patients with Moyamoya disease (MMD), in whom impaired cerebrovascular reactivity and limited cerebral blood flow reserve render cerebral perfusion highly susceptible to hemodynamic fluctuations.

In a retrospective study involving 1,497 MMD patients, Li et al. demonstrated that marked increases in the ARV of systolic, diastolic, and MAP were significantly associated with a higher risk of early postoperative cerebral infarction, with ORs of approximately 3–4 (27). Furthermore, the maximal declines in MAP and diastolic pressure (max declination) emerged as independent risk factors, suggesting that abrupt BP reductions may precipitate critical drops in cerebral perfusion, subsequently triggering extensive infarction. The investigators further emphasized that BP instability per se, rather than absolute hypotension, exerts a more deleterious effect on cerebral hemodynamics. Hence, maintaining hemodynamic stability may be of greater clinical relevance than adherence to rigid upper or lower BP thresholds. A similar pattern has been observed in pediatric populations. Zhu and He et al. analyzed 444 hemispheres from 296 pediatric MMD patients and found that ARV-MAP and maximal MAP decline were both strongly correlated with postoperative TIA progression (OR = 4.731 and 1.271, respectively; P < 0.01) (29). These findings suggest that pronounced intraoperative BP fluctuations or abrupt decreases can compromise regional cerebral perfusion, leading to ischemic injury. Given that children possess less mature autoregulatory capacity and are more dependent on hyperperfusion to sustain cerebral oxygen delivery, intraoperative BP instability poses an even greater risk of both functional and structural ischemia. Consequently, maintaining stable hemodynamics during pediatric revascularization is crucial for minimizing postoperative stroke or TIA events. Consistent conclusions were drawn by Valencia Morales et al. in a case-control study, which revealed that patients exhibiting greater intraoperative BP variance not only faced an elevated risk of early postoperative infarction but also tended to develop more extensive lesions. Notably, large territorial infarctions were more often linked to abrupt BP declines, whereas regional infarctions were predominantly associated with sustained oscillations (27). Collectively, these findings indicate that dynamic stability of intraoperative BP may better reflect the equilibrium of cerebral perfusion than its absolute level. For patients with compromised vascular reserve, individualized and continuous BP management strategies should be prioritized over fixed numerical targets.

In summary, excessive IBPV—whether through rapid fluctuations or sudden declines—substantially increases the risk of postoperative cerebral ischemia (Table 5). From a pathophysiological perspective, frequent fluctuations in blood pressure can repeatedly alter shear stress on the vascular wall, imposing mechanical strain and functional injury to the endothelium. This process is believed to promote the formation and destabilization of atherosclerotic plaques, thereby increasing the risk of intravascular thrombosis and ischemic events (32). Maintaining a relatively stable BP trajectory and avoiding transient, sharp deviations, particularly during revascularization or anesthetic transitions, represent key strategies for stroke prevention. Future investigations should integrate cerebral blood flow monitoring with hemodynamic indices to establish individualized BP management models centered on variability control, thereby reducing the incidence of perioperative stroke and improving postoperative neurological outcomes.