Association between age and angiographic vasospasm in mechanically ventilated patients with aneurysmal subarachnoid hemorrhage: a secondary analysis Authors: Jianhui Huang, Cunrong Chen, Tianlai Lin, Zhirong Ding, Rongda Cai, Youli Chen Published: 2025 ABSTRACT BackgroundAge is defined as the patient’s chronological age (in years) at hospital admission. The association between age and angiographic vasospasm remains controversial. This study aimed to explore the association between age and angiographic vasospasm in mechanically ventilated patients with aneurysmal subarachnoid hemorrhage (aSAH).MethodsA secondary analysis was performed based on a retrospective study conducted from 2010 to 2015 in the intensive care unit (ICU) of a French university hospital. Factors associated with angiographic vasospasm were identified using univariate and multivariate logistic regression analyses. The non-linear relationship between age and angiographic vasospasm was assessed using restricted cubic spline regression, and the threshold effect was evaluated using a two-piecewise linear regression model.ResultsThe study included 231 mechanically ventilated patients with aSAH. Their median age was 56 years (range, 19–84 years). Patients were divided into two groups based on the median age of 56 years: younger (< 56 years) and older (≥ 56 years). The rate of angiographic vasospasm was 34.63% (N = 80). After adjusting for potential confounders, patients in the older group were less likely to develop angiographic vasospasm [odds ratio (OR), 0.40; 95% confidence interval (CI), 0.19–0.85; p = 0.017] compared with those in the younger group. After adjusting for confounders, when age was ≤ 53 years, the risk of angiographic vasospasm increased by 13% for each year increase in age (OR, 1.13; 95% CI, 1.05–1.22; p = 0.002). For age > 53 years, the risk of angiographic vasospasm decreased by 17% for each year increase in age (OR, 0.83; 95% CI, 0.77–0.89; p < 0.001).ConclusionAge was non-linearly associated with the risk of angiographic vasospasm in mechanically ventilated patients with aSAH. Background The global incidence of subarachnoid hemorrhage is estimated to be 6.1 cases per 100,000 person-years [1, 2]. It has been reported that 38.5% of patients with aneurysmal subarachnoid hemorrhage (aSAH) require mechanical ventilation, with an in-hospital mortality rate as high as 54.6% [3]. Angiographic vasospasm is a serious but common complication after aSAH [4]. Previous studies have produced inconsistent results regarding the association between age and the risk of angiographic vasospasm. In this context, age is defined as the patient’s chronological age (in years) at hospital admission. Some studies have reported that older age is associated with a lower risk of angiographic vasospasm [5, 6] and that older age (e.g., > 65 years) has a protective effect [7, 8]. However, other studies have found no association between age and angiographic vasospasm [9, 10]. Therefore, the relationship remains controversial. This study aimed to evaluate the association between age and angiographic vasospasm in mechanically ventilated patients with aSAH. For patients in age groups at high risk for angiographic vasospasm, neurologic examinations and transcranial Doppler monitoring could be appropriately intensified to identify those who may require confirmatory imaging [11, 12]. Early detection and intervention for angiographic vasospasm may help improve patient prognosis [13]. Methods Study design and participants This secondary analysis used data from a retrospective study conducted in the intensive care unit (ICU) of a French university hospital between 2010 and 2015 [14]. A total of 236 patients with aSAH requiring mechanical ventilation were initially included. 5 patients with missing data on the duration of mechanical ventilation were excluded, resulting in a final cohort of 231. The research protocol (Q-2015–09-07) and the relevant ethical issues for the original study were published previously, and these were waived for this secondary analysis [14]. The study was carried out in accordance with the ethical standards of the Declaration of Helsinki. Data were obtained from the Dryad Digital Repository (10.5061/dryad.47d7wm3b4). The dataset is made available under the Creative Commons Zero (CC0) 1.0 Universal Public Domain Dedication, which explicitly permits unrestricted reuse, reanalysis, and dissemination for any purpose. The original study explored risk factors for poor 1-year outcome (survival and degree of disability) [14], whereas our secondary analysis investigates the association between age and angiographic vasospasm. Data collection The data included sex, age (in years) at hospital admission, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, World Federation of Neurosurgical Societies (WFNS) score, Fisher’s score, Simplified Acute Physiology Score II (SAPS II), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring external ventricular drain (EVD), rebleeding, angiographic vasospasm, length of ICU stay, and withdrawal of care. SAPS II scores were dichotomized at the median. Angiographic vasospasm was defined as a reduction of ≥ 1/3 in arterial diameter, as assessed by computed tomography angiography or digital subtraction angiography [14]. Angiographic vasospasm could occur in all intracranial arteries, but most commonly affects the internal carotid artery, middle cerebral artery, and anterior communicating artery on the side ipsilateral to the ruptured aneurysm [11, 15, 16]. Transcranial Doppler monitoring was performed daily to detect ultrasonographic vasospasm [14]. If ultrasonographic vasospasm was suspected, computed tomography angiography or digital subtraction angiography was performed to confirm angiographic vasospasm [14]. Statistical analysis Patients were divided into two groups based on the median age of 56 years: younger (< 56 years) and older (≥ 56 years). The median split is widely used in research as a nonparametric, data-driven approach that avoids bias introduced by selecting arbitrary thresholds [17, 18]. Categorical variables were presented as frequencies and percentages (n, %). For categorical variables, the chi-squared test or Fisher’s exact test was used. Univariate and multivariate logistic regression analyses were conducted to explore the association between age and angiographic vasospasm. Stratified logistic regression was used to examine the effect of age on angiographic vasospasm across different strata. Restricted cubic spline regression was used to assess the non-linear association between age and angiographic vasospasm. A two-piecewise logistic regression model was then applied to identify the threshold effect along the fitted curve [19, 20]. The inflection point was calculated automatically by the recursive method using the maximum-likelihood model [19, 20]. A log-likelihood ratio test (LRT) was also performed, and the one-line linear regression model was compared with the two-piecewise linear model [19, 20]. All statistical analyses were performed with EmpowerStats (www.empowerstats.com, X&Y solutions, Boston, Massachusetts, USA) and R software version 3.6.1 (http://www.r-project.org). A two-sided p < 0.05 was considered statistically significant. Results Patient and baseline characteristics The analysis included 231 mechanically ventilated patients with aSAH. Baseline demographic and clinical features by age were summarized in Table 1. Overall, 38.10% of patients were male. Angiographic vasospasm occurred in 80 patients (34.63%). The two age groups differed significantly at baseline, including in sex, tobacco use, cardiovascular disease, and withdrawal of care. The incidence of angiographic vasospasm was higher in the younger group than in the older group (42.34% vs. 27.50%, p = 0.016).Table 1The baseline characteristics of different groupsVariableTotalYounger (< 56 years)Older (≥ 56 years)p-valueNumber231111120Male88 (38.10%)35 (31.53%)53 (44.17%)0.048Tobacco use75 (32.47%)47 (42.34%)28 (23.33%)0.002Alcohol abuse20 (8.66%)9 (8.11%)11 (9.17%)0.775Diabetes8 (3.46%)3 (2.70%)5 (4.17%)0.723Cardiovascular disease40 (17.32%)11 (9.91%)29 (24.17%)0.004Aneurysm location0.803Anterior193 (83.91%)93 (84.55%)100 (83.33%)Posterior37 (16.09%)17 (15.45%)20 (16.67%)Glasgow Coma Scale0.251 ≤ 8160 (75.12%)73 (71.57%)87 (78.38%) > 853 (24.88%)29 (28.43%)24 (21.62%)WFNS score0.330Ⅲ—Ⅳ118 (51.08%)53 (47.75%)65 (54.17%)Ⅴ113 (48.92%)58 (52.25%)55 (45.83%)Fisher’s score0.743Ⅰ—Ⅲ50 (21.65%)23 (20.72%)27 (22.50%)Ⅳ181 (78.35%)88 (79.28%)93 (77.50%)SAPS II0.058 < 43112 (48.48%)61 (54.95%)51 (42.50%) ≥ 43119 (51.52%)50 (45.05%)69 (57.50%)Aneurysm-securing delay0.455 < 12 h130 (56.28%)67 (60.36%)63 (52.50%) ≥ 12 h, < 72 h81 (35.06%)36 (32.43%)45 (37.50%)Others20 (8.66%)8 (7.21%)12 (10%)Aneurysm-securing procedure0.801Coiling146 (63.20%)68 (61.26%)78 (65%)Clipping62 (26.84%)32 (28.83%)30 (25%)Others23 (9.96%)11 (9.91%)12 (10%)Intracerebral hemorrhage108 (46.75%)55 (49.55%)53 (44.17%)0.413Early brain injuries165 (71.74%)79 (71.17%)86 (72.27%)0.853Hydrocephalus requiring EVD154 (66.67%)73 (65.77%)81 (67.50%)0.780Rebleeding33 (14.29%)14 (12.61%)19 (15.83%)0.485Length of ICU stay (days)0.702 ≤ 439 (16.88%)19 (17.12%)20 (16.67%) > 4, ≤ 1462 (26.84%)27 (24.32%)35 (29.17%) > 14130 (56.28%)65 (58.56%)65 (54.17%)Withdrawal of care0.020No210 (90.91%)106 (95.50%)104 (86.67%)Yes21 (9.09%)5 (4.50%)16 (13.33%)Angiographic vasospasm80 (34.63%)47 (42.34%)33 (27.50%)0.018Date were presented as n (%)WFNS World Federation of Neurosurgical Societies, SAPS II Simplified Acute Physiology Score II, EVD External ventricular drain, ICU Intensive care unitData on the Glasgow Coma Scale were missing for 18 patients. Data on the aneurysm localization and early brain injuries were missing for 1 patients Table 1The baseline characteristics of different groupsVariableTotalYounger (< 56 years)Older (≥ 56 years)p-valueNumber231111120Male88 (38.10%)35 (31.53%)53 (44.17%)0.048Tobacco use75 (32.47%)47 (42.34%)28 (23.33%)0.002Alcohol abuse20 (8.66%)9 (8.11%)11 (9.17%)0.775Diabetes8 (3.46%)3 (2.70%)5 (4.17%)0.723Cardiovascular disease40 (17.32%)11 (9.91%)29 (24.17%)0.004Aneurysm location0.803Anterior193 (83.91%)93 (84.55%)100 (83.33%)Posterior37 (16.09%)17 (15.45%)20 (16.67%)Glasgow Coma Scale0.251 ≤ 8160 (75.12%)73 (71.57%)87 (78.38%) > 853 (24.88%)29 (28.43%)24 (21.62%)WFNS score0.330Ⅲ—Ⅳ118 (51.08%)53 (47.75%)65 (54.17%)Ⅴ113 (48.92%)58 (52.25%)55 (45.83%)Fisher’s score0.743Ⅰ—Ⅲ50 (21.65%)23 (20.72%)27 (22.50%)Ⅳ181 (78.35%)88 (79.28%)93 (77.50%)SAPS II0.058 < 43112 (48.48%)61 (54.95%)51 (42.50%) ≥ 43119 (51.52%)50 (45.05%)69 (57.50%)Aneurysm-securing delay0.455 < 12 h130 (56.28%)67 (60.36%)63 (52.50%) ≥ 12 h, < 72 h81 (35.06%)36 (32.43%)45 (37.50%)Others20 (8.66%)8 (7.21%)12 (10%)Aneurysm-securing procedure0.801Coiling146 (63.20%)68 (61.26%)78 (65%)Clipping62 (26.84%)32 (28.83%)30 (25%)Others23 (9.96%)11 (9.91%)12 (10%)Intracerebral hemorrhage108 (46.75%)55 (49.55%)53 (44.17%)0.413Early brain injuries165 (71.74%)79 (71.17%)86 (72.27%)0.853Hydrocephalus requiring EVD154 (66.67%)73 (65.77%)81 (67.50%)0.780Rebleeding33 (14.29%)14 (12.61%)19 (15.83%)0.485Length of ICU stay (days)0.702 ≤ 439 (16.88%)19 (17.12%)20 (16.67%) > 4, ≤ 1462 (26.84%)27 (24.32%)35 (29.17%) > 14130 (56.28%)65 (58.56%)65 (54.17%)Withdrawal of care0.020No210 (90.91%)106 (95.50%)104 (86.67%)Yes21 (9.09%)5 (4.50%)16 (13.33%)Angiographic vasospasm80 (34.63%)47 (42.34%)33 (27.50%)0.018Date were presented as n (%)WFNS World Federation of Neurosurgical Societies, SAPS II Simplified Acute Physiology Score II, EVD External ventricular drain, ICU Intensive care unitData on the Glasgow Coma Scale were missing for 18 patients. Data on the aneurysm localization and early brain injuries were missing for 1 patients The baseline characteristics of different groups Date were presented as n (%) WFNS World Federation of Neurosurgical Societies, SAPS II Simplified Acute Physiology Score II, EVD External ventricular drain, ICU Intensive care unit Data on the Glasgow Coma Scale were missing for 18 patients. Data on the aneurysm localization and early brain injuries were missing for 1 patients Univariate and multivariate analysis Univariate analysis of angiographic vasospasm among mechanically ventilated patients with aSAH was presented in Table 2. The WFNS score, SAPS II score, aneurysm-securing delay, aneurysm-securing procedure, early brain injury, hydrocephalus requiring EVD, and length of ICU stay were associated with angiographic vasospasm. Variables with a p < 0.05 in univariate analysis, along with other clinically relevant factors, were included in the multivariate logistic regression.Table 2Univariate analysis of angiographic vasospasmVariableOR (95% CI)p-valueSex0.892Female1.0Male0.96 (0.55, 1.68)Tobacco use0.372No1.0Yes1.30 (0.73, 2.30)Alcohol abuse0.051No1.0Yes2.52 (1.00, 6.35)Diabetes0.862No1.0Yes1.14 (0.26, 4.89)Cardiovascular disease0.080No1.0Yes0.49 (0.22, 1.09)Aneurysm location0.282Anterior1.0Posterior0.65 (0.30, 1.42)Glasgow Coma Scale0.152 ≤ 81.0 > 81.59 (0.84, 3.01)WFNS score0.025Ⅲ—Ⅳ1.0Ⅴ0.53 (0.31, 0.92)Fisher's Score0.368Ⅰ—Ⅲ1.0Ⅳ0.74 (0.39, 1.42)SAPS II0.047 < 431.0 ≥ 430.57 (0.33, 0.99)Aneurysm-securing delay < 12 h1.012 h—72 h1.09 (0.62, 1.94)0.757Others0.20 (0.04, 0.88)0.034Aneurysm-securing procedureCoiling1.0Clipping0.27 (0.13, 0.56) < 0.001Others0.26 (0.09, 0.81)0.020Intracerebral hemorrhage0.346No1.0Yes0.77 (0.45, 1.33)Early brain injury0.019No1.0Yes0.49 (0.27, 0.89)Hydrocephalus requiring EVD0.005No1.0Yes2.44 (1.31, 4.57)Rebleeding0.339No1.0Yes0.67 (0.30, 1.52)Length of ICU stay (days) ≤ 41.0 > 4, ≤ 148.17 (1.78, 37.44)0.007 > 1415.37 (3.56, 66.47) < 0.001Withdrawal of careNo1.0Yes0.56 (0.20, 1.60)0.280OR Odds ratio, CI Confidence interval, WFNS World Federation of Neurosurgical Societies, SAPS II Simplified Acute Physiology Score Ⅱ, EVD External ventricular drain, ICU Intensive care unit Table 2Univariate analysis of angiographic vasospasmVariableOR (95% CI)p-valueSex0.892Female1.0Male0.96 (0.55, 1.68)Tobacco use0.372No1.0Yes1.30 (0.73, 2.30)Alcohol abuse0.051No1.0Yes2.52 (1.00, 6.35)Diabetes0.862No1.0Yes1.14 (0.26, 4.89)Cardiovascular disease0.080No1.0Yes0.49 (0.22, 1.09)Aneurysm location0.282Anterior1.0Posterior0.65 (0.30, 1.42)Glasgow Coma Scale0.152 ≤ 81.0 > 81.59 (0.84, 3.01)WFNS score0.025Ⅲ—Ⅳ1.0Ⅴ0.53 (0.31, 0.92)Fisher's Score0.368Ⅰ—Ⅲ1.0Ⅳ0.74 (0.39, 1.42)SAPS II0.047 < 431.0 ≥ 430.57 (0.33, 0.99)Aneurysm-securing delay < 12 h1.012 h—72 h1.09 (0.62, 1.94)0.757Others0.20 (0.04, 0.88)0.034Aneurysm-securing procedureCoiling1.0Clipping0.27 (0.13, 0.56) < 0.001Others0.26 (0.09, 0.81)0.020Intracerebral hemorrhage0.346No1.0Yes0.77 (0.45, 1.33)Early brain injury0.019No1.0Yes0.49 (0.27, 0.89)Hydrocephalus requiring EVD0.005No1.0Yes2.44 (1.31, 4.57)Rebleeding0.339No1.0Yes0.67 (0.30, 1.52)Length of ICU stay (days) ≤ 41.0 > 4, ≤ 148.17 (1.78, 37.44)0.007 > 1415.37 (3.56, 66.47) < 0.001Withdrawal of careNo1.0Yes0.56 (0.20, 1.60)0.280OR Odds ratio, CI Confidence interval, WFNS World Federation of Neurosurgical Societies, SAPS II Simplified Acute Physiology Score Ⅱ, EVD External ventricular drain, ICU Intensive care unit Univariate analysis of angiographic vasospasm OR Odds ratio, CI Confidence interval, WFNS World Federation of Neurosurgical Societies, SAPS II Simplified Acute Physiology Score Ⅱ, EVD External ventricular drain, ICU Intensive care unit As shown in Table 3, the risk of angiographic vasospasm was lower in the older group [odds ratio (OR), 0.52; 95% confidence interval (CI), 0.30–0.90; p = 0.019] compared with the younger group. After adjusting for potential confounders, the risk of angiographic vasospasm remained lower in the older group (OR, 0.40; 95% CI, 0.19–0.85; p = 0.017) compared with the younger group. In subgroup analyses (Table 4), the risk of angiographic vasospasm was lower in the older group across most strata compared with the younger group. Stratified analyses were not performed in strata where the number of angiographic vasospasm events was < 5.Table 3Association between age and angiographic vasospasmVariableUnadjustedModel IModel ⅡOR (95% CI)p-valueOR (95% CI)p-valueOR (95% CI)p-valueYounger (< 56 years)1.01.01.0Older (≥ 56 years)0.52 (0.30, 0.90)0.0190.45 (0.22, 0.93)0.0300.40 (0.19, 0.85)0.017OR Odds ratio, CI Confidence intervalModel I adjusted for: sex, tobacco use, alcohol abuse, cardiovascular disease, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, early brain injury, hydrocephalus requiring EVD, length of ICU stay, and withdrawal of careModel II adjust for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), Fisher’s score (Ⅰ—Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleeding, length of ICU stay, and withdrawal of careTable 4Stratified analysis of angiographic vasospasmVariableTotalEventOR (95% CI)Younger (< 56 years)Older (≥ 56 years)Sex Female143501.00.65 (0.33, 1.31) Male88301.00.34 (0.14, 0.86)Tobacco use No156511.00.48 (0.24, 0.95) Yes75291.00.64 (0.24, 1.71)Alcohol abuse No211691.00.47 (0.26, 0.85) Yes20111.00.96 (0.16, 5.64)Diabetes No223771.00.54 (0.31, 0.94) Yes83--Cardiovascular disease No191711.00.54 (0.30, 0.98) Yes4091.00.70 (0.14, 3.45)Aneurysm location Anterior193701.00.57 (0.31, 1.03) Posterior37101.00.25 (0.05, 1.20)Glasgow Coma Scale ≤ 8160521.00.69 (0.35, 1.34) > 853231.00.33 (0.11, 1.05)WFNS score III—IV118491.00.49 (0.23, 1.04) Ⅴ113311.00.48 (0.20, 1.12)Fisher’s Score Ⅰ—Ⅲ50201.00.55 (0.17, 1.71) Ⅳ181601.00.50 (0.27, 0.94)SAPS II < 43112461.00.47 (0.22, 1.03) ≥ 43119341.00.63 (0.28, 1.41)Aneurysm-securing delay < 12 h130471.00.90 (0.44, 1.85) ≥ 12 h, < 72 h81311.00.26 (0.10, 0.67) Others202--Aneurysm-securing procedure Coiling146651.00.47 (0.24, 0.91) Clipping62111.00.87 (0.23, 3.20) Others234--Intracerebral hemorrhage No123461.00.49 (0.23, 1.03) Yes108341.00.53 (0.23, 1.21)Early brain injuries No65301.00.57 (0.21, 1.54) Yes165491.00.46 (0.23, 0.92)Hydrocephalus requiring EVD No77171.00.61 (0.21, 1.82) Yes154631.00.46 (0.24, 0.88)Rebleeding No198711.00.53 (0.29, 0.95) Yes3391.00.48 (0.10, 2.27)Length of ICU survival (days) ≤ 4392-- > 4, ≤ 1462191.01.09 (0.37, 3.24) > 14130591.00.39 (0.19, 0.79)Withdrawal of care No210751.00.60 (0.34, 1.06) Yes2151.00.10 (0.01, 0.97)OR Odds ratio, CI Confidence interval, WFNS World Federation of Neurosurgical Societies, SAPS II Simplified Acute Physiology Score II, EVD External ventricular drain, ICU, Intensive care unit Table 3Association between age and angiographic vasospasmVariableUnadjustedModel IModel ⅡOR (95% CI)p-valueOR (95% CI)p-valueOR (95% CI)p-valueYounger (< 56 years)1.01.01.0Older (≥ 56 years)0.52 (0.30, 0.90)0.0190.45 (0.22, 0.93)0.0300.40 (0.19, 0.85)0.017OR Odds ratio, CI Confidence intervalModel I adjusted for: sex, tobacco use, alcohol abuse, cardiovascular disease, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, early brain injury, hydrocephalus requiring EVD, length of ICU stay, and withdrawal of careModel II adjust for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), Fisher’s score (Ⅰ—Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleeding, length of ICU stay, and withdrawal of care Association between age and angiographic vasospasm OR Odds ratio, CI Confidence interval Model I adjusted for: sex, tobacco use, alcohol abuse, cardiovascular disease, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, early brain injury, hydrocephalus requiring EVD, length of ICU stay, and withdrawal of care Model II adjust for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), Fisher’s score (Ⅰ—Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleeding, length of ICU stay, and withdrawal of care Table 4Stratified analysis of angiographic vasospasmVariableTotalEventOR (95% CI)Younger (< 56 years)Older (≥ 56 years)Sex Female143501.00.65 (0.33, 1.31) Male88301.00.34 (0.14, 0.86)Tobacco use No156511.00.48 (0.24, 0.95) Yes75291.00.64 (0.24, 1.71)Alcohol abuse No211691.00.47 (0.26, 0.85) Yes20111.00.96 (0.16, 5.64)Diabetes No223771.00.54 (0.31, 0.94) Yes83--Cardiovascular disease No191711.00.54 (0.30, 0.98) Yes4091.00.70 (0.14, 3.45)Aneurysm location Anterior193701.00.57 (0.31, 1.03) Posterior37101.00.25 (0.05, 1.20)Glasgow Coma Scale ≤ 8160521.00.69 (0.35, 1.34) > 853231.00.33 (0.11, 1.05)WFNS score III—IV118491.00.49 (0.23, 1.04) Ⅴ113311.00.48 (0.20, 1.12)Fisher’s Score Ⅰ—Ⅲ50201.00.55 (0.17, 1.71) Ⅳ181601.00.50 (0.27, 0.94)SAPS II < 43112461.00.47 (0.22, 1.03) ≥ 43119341.00.63 (0.28, 1.41)Aneurysm-securing delay < 12 h130471.00.90 (0.44, 1.85) ≥ 12 h, < 72 h81311.00.26 (0.10, 0.67) Others202--Aneurysm-securing procedure Coiling146651.00.47 (0.24, 0.91) Clipping62111.00.87 (0.23, 3.20) Others234--Intracerebral hemorrhage No123461.00.49 (0.23, 1.03) Yes108341.00.53 (0.23, 1.21)Early brain injuries No65301.00.57 (0.21, 1.54) Yes165491.00.46 (0.23, 0.92)Hydrocephalus requiring EVD No77171.00.61 (0.21, 1.82) Yes154631.00.46 (0.24, 0.88)Rebleeding No198711.00.53 (0.29, 0.95) Yes3391.00.48 (0.10, 2.27)Length of ICU survival (days) ≤ 4392-- > 4, ≤ 1462191.01.09 (0.37, 3.24) > 14130591.00.39 (0.19, 0.79)Withdrawal of care No210751.00.60 (0.34, 1.06) Yes2151.00.10 (0.01, 0.97)OR Odds ratio, CI Confidence interval, WFNS World Federation of Neurosurgical Societies, SAPS II Simplified Acute Physiology Score II, EVD External ventricular drain, ICU, Intensive care unit Stratified analysis of angiographic vasospasm OR Odds ratio, CI Confidence interval, WFNS World Federation of Neurosurgical Societies, SAPS II Simplified Acute Physiology Score II, EVD External ventricular drain, ICU, Intensive care unit Non-linear association and threshold effect of age on angiographic vasospasm In both unadjusted and adjusted models, age showed a non-linear association with angiographic vasospasm (Fig. 1). After adjusting for confounders, the risk of angiographic vasospasm increased with age and then gradually declined, with the highest risk observed at 53 years. As shown in Table 5, before the age of 53 years, the risk increased by 13% for each additional year (OR, 1.13; 95% CI, 1.05–1.22). After the age of 53 years, the risk decreased by 17% for each additional year (OR, 0.83; 95% CI, 0.77–0.89). Angiographic vasospasm often occurred between days 4 and 14 after aSAH onset [6]. After excluding patients who underwent withdrawal of care, the non-linear association between age and angiographic vasospasm persisted among patients with an ICU stay > 4 days or > 14 days, as shown in Fig. 2 and Table 6.Fig. 1Smoothed association between age and angiographic vasospasm. Adjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ-Ⅳ, Ⅴ), Fisher’s score (Ⅰ-Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleeding, length of ICU stay (days), and withdrawal of careTable 5Threshold effect analysis of age on angiographic vasospasm using two-piecewise logistic regression modelModelsUnadjusted OR (95% CI)Adjusted OR (95% CI)Model Ⅰ One line effect0.98 (0.96, 1.0)0.96 (0.93, 1.0)Model Ⅱ Turning point (K)5253 Age ≤ K1.09 (1.03, 1.15)1.13 (1.05, 1.22) Age > K0.90 (0.86, 0.95)0.83 (0.77, 0.89) p-value for LRT test < 0.001 < 0.001OR Odds ratio, CI Confidence interval, LRT Log-likelihood ratio testAdjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), Fisher’s score (Ⅰ—Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleeding, length of ICU stay, and withdrawal of careFig. 2Smoothed association between age and angiographic vasospasm. Adjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ-Ⅳ, Ⅴ), Fisher’s score (Ⅰ-Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleedingTable 6Threshold effect analysis of age on angiographic vasospasm using two-piecewise logistic regression modelModelsUnadjusted OR (95% CI)Adjusted OR (95% CI)Length of ICU stay > 4 daysModel Ⅰ One line effect0.98 (0.96, 1.01)0.98 (0.95, 1.02)Model Ⅱ Turning point (K)5252 Age ≤ K1.08 (1.02, 1.16)1.15 (1.05, 1.26) Age > K0.91 (0.87, 0.96)0.86 (0.80, 0.93) p-value for LRT test < 0.001 < 0.001Length of ICU stay > 14 daysModel Ⅰ One line effect0.98 (0.95, 1.01)0.96 (0.91, 1.01)Model Ⅱ Turning point (K)5253 Age ≤ K1.08 (1.00, 1.17)1.12 (0.99, 1.26) Age > K0.91 (0.85, 0.96)0.84 (0.75, 0.93) p-value for LRT test0.003 < 0.001OR Odds ratio, CI Confidence interval, LRT Log-likelihood ratio testAdjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), Fisher’s score (Ⅰ—Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injuries, hydrocephalus requiring EVD, and rebleeding Smoothed association between age and angiographic vasospasm. Adjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ-Ⅳ, Ⅴ), Fisher’s score (Ⅰ-Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleeding, length of ICU stay (days), and withdrawal of care Table 5Threshold effect analysis of age on angiographic vasospasm using two-piecewise logistic regression modelModelsUnadjusted OR (95% CI)Adjusted OR (95% CI)Model Ⅰ One line effect0.98 (0.96, 1.0)0.96 (0.93, 1.0)Model Ⅱ Turning point (K)5253 Age ≤ K1.09 (1.03, 1.15)1.13 (1.05, 1.22) Age > K0.90 (0.86, 0.95)0.83 (0.77, 0.89) p-value for LRT test < 0.001 < 0.001OR Odds ratio, CI Confidence interval, LRT Log-likelihood ratio testAdjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), Fisher’s score (Ⅰ—Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleeding, length of ICU stay, and withdrawal of care Threshold effect analysis of age on angiographic vasospasm using two-piecewise logistic regression model OR Odds ratio, CI Confidence interval, LRT Log-likelihood ratio test Adjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), Fisher’s score (Ⅰ—Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleeding, length of ICU stay, and withdrawal of care Smoothed association between age and angiographic vasospasm. Adjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ-Ⅳ, Ⅴ), Fisher’s score (Ⅰ-Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injury, hydrocephalus requiring EVD, rebleeding Table 6Threshold effect analysis of age on angiographic vasospasm using two-piecewise logistic regression modelModelsUnadjusted OR (95% CI)Adjusted OR (95% CI)Length of ICU stay > 4 daysModel Ⅰ One line effect0.98 (0.96, 1.01)0.98 (0.95, 1.02)Model Ⅱ Turning point (K)5252 Age ≤ K1.08 (1.02, 1.16)1.15 (1.05, 1.26) Age > K0.91 (0.87, 0.96)0.86 (0.80, 0.93) p-value for LRT test < 0.001 < 0.001Length of ICU stay > 14 daysModel Ⅰ One line effect0.98 (0.95, 1.01)0.96 (0.91, 1.01)Model Ⅱ Turning point (K)5253 Age ≤ K1.08 (1.00, 1.17)1.12 (0.99, 1.26) Age > K0.91 (0.85, 0.96)0.84 (0.75, 0.93) p-value for LRT test0.003 < 0.001OR Odds ratio, CI Confidence interval, LRT Log-likelihood ratio testAdjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), Fisher’s score (Ⅰ—Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injuries, hydrocephalus requiring EVD, and rebleeding Threshold effect analysis of age on angiographic vasospasm using two-piecewise logistic regression model OR Odds ratio, CI Confidence interval, LRT Log-likelihood ratio test Adjusted for: sex, tobacco use, alcohol abuse, diabetes, cardiovascular disease, aneurysm location, Glasgow Coma Scale, WFNS score (Ⅲ—Ⅳ, Ⅴ), Fisher’s score (Ⅰ—Ⅲ, Ⅳ), SAPS II (< 43, ≥ 43), aneurysm-securing delay, aneurysm-securing procedure, intracerebral hemorrhage, early brain injuries, hydrocephalus requiring EVD, and rebleeding Discussion In this study, the association between age and angiographic vasospasm in mechanically ventilated patients with aSAH was analyzed. The incidence of angiographic vasospasm was 34.63% (80 patients), consistent with previous studies [5, 6, 10, 21]. Risk factors for angiographic vasospasm were identified, including SAPS II, aneurysm-securing delay, aneurysm-securing procedure, early brain injury, hydrocephalus requiring EVD, and length of ICU stay; these findings were consistent with previous studies [22–25]. Among mechanically ventilated patients with aSAH, older patients were less likely to develop angiographic vasospasm. Furthermore, a non-linear association between age and the risk of angiographic vasospasm was observed. For patients aged ≤ 53 years, the risk of angiographic vasospasm increased with each additional year. For patients aged > 53 years, the risk decreased with each additional year. The pathophysiological mechanisms remain unclear. In relatively young aSAH patients requiring mechanical ventilated, the risk of angiographic vasospasm increases with age, may reflect age-related declines in cerebrovascular autoregulation [26]. Additionally, this may be associated with the observation that the benefits of nimodipine in preventing cerebral vasospasm are greater in younger patients [27]. In relatively older patients, the risk of angiographic vasospasm may decrease with age. Potentially due to age-associated arteriosclerosis and an attenuated inflammatory response that can attenuate vascular reactivity to constrictive stimuli after aSAH [8]. Age-related arteriosclerosis may impair the contractility and elasticity of small arteries and arteriolar branches [28]. Vascular responsiveness to inflammatory mediators may also decrease with advancing age [29]. Angiographic vasospasm was defined as the narrowing of intracranial arteries [11, 15, 16]. It could occur in all intracranial arteries, but most commonly in the internal carotid artery, middle cerebral artery, and anterior communicating artery on the side ipsilateral to the ruptured aneurysm [10, 30]. Since this study was a secondary analysis, it was unclear whether the original authors assessed all intracranial arteries or only the specific vascular territories associated with the ruptured aneurysm. This study has several limitations. First, it was a single-center retrospective cohort study, which may limit the generalizability of the findings to broader populations. Second, because the secondary analysis was restricted to patients with aSAH who required mechanical ventilation, the findings may not be generalizable to patients managed without mechanical ventilation. The findings still offer a reference for the association between age and angiographic vasospasm in mechanically ventilated patients. Third, the very elderly subgroup was small (only 4 patients aged ≥ 80 years), which limited the applicability to this population. Fourth, despite adjustment for potential confounders, unmeasured variables may have influenced the results. Fifth, owing to limitations inherent in the primary dataset, the association between the modified Fisher’s score and angiographic vasospasm could not be assessed. Conclusion Age was non-linearly associated with angiographic vasospasm in mechanically ventilated patients with aSAH.