Sex Disparities in Cardiogenic Shock: Risk Factors, Treatment Intensity, and Mortality in a Single Latin American Country

Study design and setting

We analyzed a coronary care unit database (n = 28,054) from late 2015 to September 2023 at the Instituto Nacional de Cardiología Ignacio Chavez, which is an academic tertiary and reference center in Mexico City with a focus on only cardiovascular diseases from which we include patients with the diagnosis of CS was defined as Society for Cardiac Angiography and Interventions-Cardiogenic Shock Working Group (SCAI-CSWG) score (1), a sustained episode of at least 1 of the following: systolic blood pressure (SBP) <90 mmHg for at least 30 min, use of vasoactive agents to maintain SBP, cardiac index <2.2 L/min/m² in the absence of hypovolemia, each determined to be secondary to cardiac dysfunction, or use of an MCS device for clinically suspected CS, and include those in SCAI B or higher maximum score at the first 72-h. No additional exclusion criteria were applied, except for patients without a confirmed diagnosis of cardiogenic shock.

The Research and Ethics Committee approved the study protocol; patient consent was waived because the study was retrospective and non-interventional. All procedures were conducted according to the Declaration of Helsinki and local regulations.

Outcomes

The primary outcome was in-hospital mortality. Secondary analyses focused on: (1) identifying independent predictors of mortality within each subgroup (AMI-CS and non-AMI-CS, stratified by sex); and (2) evaluating whether sex was independently associated with mortality even after adjustment for demographics, comorbidities, treatment intensity, and other clinically relevant covariates.

AMI cardiogenic shock

Women presented at an older age (67 vs. 60 years, P < 0.001), while men had higher BMI (P < 0.001). Hypertension and diabetes were more common in women (P < 0.001), whereas COPD and prior MI predominated in men (P = 0.005 for both). Women more often presented with NSTEMI (P < 0.001).

Cardiac and laboratory findings also differed. Women had slightly higher LVEF, but lower hemoglobin, estimated glomerular filtration (eGFR) by CKD-EPI, sodium, and chloride, with higher glucose (P for all <0.001). Liver enzymes were lower in women (AST/ALT, P < 0.001), while LDH was higher in men (P = 0.002). CRP did not differ. Minimum pH was lower in women (P < 0.001), though peak lactate was similar.

Management disparities were evident: reperfusion was less frequent in women (62% vs. 71%, P < 0.001), who also required more hemodialysis (P = 0.009) and multiple vasopressors, particularly norepinephrine (P < 0.001) and vasopressin (P = 0.006).

Additionally, MCS was more utilized in men (P = 0.011); specific use of intra-aortic balloon pump (IABP) was 14.7% in men vs. 11.6% in women (P = 0.012), but as of the date of the study, the low numbers of ECMO (8 men vs. 0 women) and Impella (2 men vs. 1 woman) were for the initiation of the utilization of these MCS in our media (P > 0.05).

Outcomes were worse in women, with higher AKI (P < 0.001), stroke (P = 0.03), and mortality (24.6% vs. 16.3%, P < 0.001). They were also more frequently classified in advanced SCAI-CSWG stages (P < 0.001; Table 1).

Non-AMI cardiogenic shock

Women presented at an older age (P < 0.001), while men had higher BMI (P = 0.012) and smoking prevalence (53% vs. 16.1%, P < 0.001). Men also had more dyslipidemia, hypertension, COPD, and prior MI, whereas women more often had a history of atrial fibrillation.

Women showed higher LVEF (54% vs. 40%, P < 0.001) but lower hemoglobin (P < 0.001) and eGFR (P = 0.001). Men had higher BUN, creatinine, and leukocyte counts, while women had higher platelet counts and glucose levels (all P < 0.001). AST and ALT were higher in men, LDH was similar, and minimum pH was lower in women (P = 0.003), though peak lactate did not differ.

Regarding management, PAC use was more common in men (2.6% vs. 1.3%, P = 0.001). Rates of mechanical ventilation and hemodialysis were similar. Vasoactive therapy showed sex-specific patterns: levosimendan (6.1% vs. 2.6%, P < 0.001) and dobutamine (P < 0.001) were used more often in men, while norepinephrine was more frequent in women (P = 0.045). MCS increased in men (1.9% vs. 1.2%, P = 0.058). No Impella was used, and ECMO use was rare and balanced (4 men and 6 women, P = 0.36).

Outcomes again favored men. Mortality was higher in women (P = 0.025), as was AKI (39.6% vs. 35.6%, P = 0.007). No significant sex difference was observed in SCAI stage distribution (Table 1).

Specific risk factors for mortality in CS in women and men

Tables 2 and 3 delineate the differences observed in survivors vs. non-survivors in men and women with AMI-CS and non-AMI-CS. Moreover, Table S4 in the supplementary data represents the univariate hazard ratio (HR) of men and women with AMI-CS and non-AMI-CS with in-hospital mortality.

Specific risk factors in the multivariate analysis of mortality in AMI-CS

In men with AMI-CS, type 2 diabetes (P = 0.003) and STEMI presentation (P = 0.004) were independently associated with higher mortality. Primary PCI was protective, whereas pharmacoinvasive and late/non-reperfusion strategies were not. PAC use was associated with lower mortality (HR 0.75, 95% CI 0.59–0.96, P = 0.023). Mechanical ventilation markedly increased the risk. Mortality rose progressively with higher SCAI-CSWG: HRs were 6.29 (95% CI 4.06–9.75), 4.07 (2.66–6.23), and 3.22 (2.12–4.90) for stages E, D, and C, respectively (all P < 0.001 vs. stage B). Additional predictors included older age, reduced LVEF, higher BUN, and elevated lactate. Albumin showed an inverse association with mortality (HR 1.01, 95% CI 1.01–1.10, P = 0.002).

In women with AMI-CS, mechanical ventilation was strongly associated with increased mortality (HR 1.76, 95% CI 1.20–2.58, P = 0.004). Higher SCAI-CSWG stage conferred a stepwise increase in risk compared with stage B: HRs were 4.90 (95% CI 2.75–8.72), 3.02 (1.63–5.59), and 2.86 (1.62–5.04) for stages E, D, and C, respectively (all P < 0.001). Additional predictors included older age, reduced LVEF, and elevated lactate.

Specific risk factors in the multivariate analysis of mortality in non-AMI-CS

In men with non-AMI-CS, higher SCAI-CSWG stages predicted mortality, with HRs of 3.71 (95% CI 2.66–5.17, P < 0.001) for stage E, 1.98 (1.34–2.84, P < 0.001) for D, and 1.61 (1.15–2.26, P = 0.006) for C compared with stage B. Mechanical ventilation and lower diastolic blood pressure were also associated with increased risk. Additional predictors included lower platelet counts, higher BUN and CRP, while higher sodium and potassium levels were protective (Figure 2 and Graphical Abstract).

Figure 2

In women with non-AMI-CS, prior heart failure and mechanical ventilation were independently associated with increased mortality. Higher SCAI-CSWG stages conferred a stepwise rise in risk: HRs were 4.24 (95% CI 3.01–5.97, P < 0.001) for stage E, 3.29 (2.24–4.84, P < 0.001) for D, and 1.62 (1.13–2.30, P = 0.008) for C compared with stage B. Additional predictors included older age, higher albumin, ALT, LDH, and CRP, whereas higher BMI was protective (Figure 2 and Graphical Abstract).

Treatment intensity in AMI and non-AMI-CS

In AMI-CS treatment intensity varied across SCAI-CSWG stages with consistent gender differences. Invasive procedures such as PAC and hemodialysis were used less frequently in women, particularly in stage E. Mechanical circulatory support – especially IABP – was more often used in men across stages D and E. Among vasoactives, norepinephrine and levosimendan were more common in men (notably stage D), whereas vasopressin use was higher in women (Table S3).

In non-AMI-CS, gender-related treatment patterns were subtler. PAC was more prevalent in men, especially in stage D, as was hemodialysis. Rates of MCS did not differ between sexes across SCAI stages. In contrast, vasoactive use diverged: norepinephrine and vasopressin were more common in women (stages C and D), while men more often received dobutamine and levosimendan across stages C–E (Table S4).

Sex-based differences in mortality and survival (AMI-CS and non–AMI-CS)

In AMI-CS, women had shorter survival (Log-rank P < 0.001), 30-day RMST of 2.11 (1.12–3.08, P < 0.001) days fewer of survival with a HR of 1.48 (1.28–1.72, P < 0.001) and adjusted HR of 1.24 (1.05–1.46; P = 0.011).

For non-AMI-CS survival was also reduced in women (Log-rank P = 0.006), a 30-day RMST, a difference of diminished survival of 0.85 (0.08–1.62, P = 0.03) days, a HR of 1.18 (1.05–1.32; P = 0.007), and a non-significant adjusted HR of 1.10 (0.96–1.25, P = 0.175; Figures 3 and 4).

Figure 3

Figure 4

Propensity score matching sex-based differences in mortality and survival (AMI-CS and non–AMI-CS)

After propensity score matching, sex-related differences in survival were attenuated. In AMI-CS (931 vs. 931), women showed a borderline difference by Log-rank testing (P = 0.048), but no significant differences were observed in 30-day RMST (1.09 days, 95% CI –0.17 to 2.35; P = 0.09) or mortality (HR 1.22, 95% CI 1.00–1.48; P = 0.052). In non-AMI-CS (1406 vs. 1406), no differences were identified in Log-rank P = 0.359, 30-day RMST 0.4 (–0.55–1.36, P = 0.407), or HR 1.07 (0.92–1.24, P = 0.366; Figures 3 and 4, full analysis supplementary Figures S3 and Table S6).

AMI-CS sex differences

In AMI-CS, women presented at an older age with more hypertension and diabetes, as in CSWG and CUPLRIT-SHOCK analysis (9, 14), in contrast to our study, since men showed a higher median BMI and smoking prevalence in this Latin cohort. Women also exhibited lower hemoglobin, lower eGFR, and higher hemo-metabolic compromise, reflected by lower 72 h-minimum pH – findings only partially captured in prior registries (CSWG, INOVA) (79).

In the realm of therapeutic interventions, sex-specific disparities were evident. Women received significantly less primary reperfusion than men, contrasting with the INOVA registry (no differences, P = 0.32) and the NIS registry (66.7% vs. 65.7%, P < 0.001), but aligning with CULPRIT-SHOCK, where women trended toward lower reperfusion rates (71415).

Regarding advanced therapies, women with AMI-CS were less likely to receive MCS and more likely to be treated with vasopressors, while no differences emerged in non-AMI-CS. Yan et al. similarly reported reduced use of LVADs (OR 0.78, 0.64–0.94) and greater vasopressor exposure (OR 1.26, 1.05–1.5) in women, without mortality differences (716).

In contrast, INOVA-SHOCK showed no sex differences, and non-AMI-CS cohorts even demonstrated higher MCS and vasopressor use in women (7).

Adverse outcomes also showed sex-specific differences. Women experienced higher rates of stroke and AKI, and prior work by Vallabhajosyula et al. demonstrated that female patients with AKI were less likely to receive advanced therapies such as MCS, mechanical ventilation, or hemodialysis, despite greater comorbidity and higher mortality (19). Consistently, in our cohort, women presented more often with advanced SCAI-CSWG (C–E) and higher unadjusted mortality, findings that contrast with the CSWG report, where admission stage distribution and outcomes were similar between sexes (9).

Non-AMI-CS sex differences

In non-AMI CS, we observed distinct sex-related patterns. Similar to CSWG and Sundermeyer et al. (918), men were younger and more likely to have prior MI, whereas women were older with a greater prevalence of atrial fibrillation and lower eGFR despite lower creatinine levels. Significantly, in our cohort, age independently predicted outcomes only in women, suggesting heightened vulnerability of older females with non-AMI CS. Echocardiographic assessment also showed higher LVEF in women, consistent with prior non-ischemic shock registries (18).

Consistent with Sundermeyer, creatinine was higher in men; we observed the same pattern. Although eGFR was not reported in CSWG or Sundermeyer, in our cohort, men had higher creatinine, but women had lower eGFR by the CKD-EPI formula. Hepatic enzymes also differed by sex: AST and ALT were higher in men, contrasting with CSWG, where ALT did not differ (women 290 vs. men 266 U/L; P = 0.07), and AST was not reported. Finally, women had lower arterial pH despite similar peak lactate, suggesting a sex-specific pathway to metabolic acidosis not captured by lactate; prior registries reported no sex differences in lactate or pH (918).

Regarding hemodynamic support, PAC use was very low overall; this contrasts sharply with CSWG, where nearly half of both sexes underwent PAC, reflecting regional differences and the legacy indication of PAC mainly for AMI-CS. Mechanical ventilation and vasoactive requirements showed no sex differences, in line with CSWG and Sundermeyer (918). The SCAI-CSWG classification demonstrated no sex disparity, in contrast to the CSWG registry, in which women appeared to have higher stages (9).

Specific sex risk factors

In AMI-CS, women exhibit a heightened reliance on creatinine levels as a prognostic marker, underscoring the importance of renal function, and possibly related to a lower intensity of therapies in this population (17). Conversely, men demonstrate a propensity toward classical risk factors such as DM2, type of MI, and the presence of primary reperfusion interventions.

Furthermore, PAC may refine risk stratification and guide therapy; however, its effect must be interpreted cautiously, given the lower use observed in non-AMI-CS and women with AMI-CS. Previously, our group demonstrated that PAC, as a guidance to decongestion, could serve as a goal in AMI-CS (1920) while Kanwar et al. reported that early PAC use in non-AMI-CS reduced mortality (1920).

Despite these sex-specific disparities, areas of convergence are evident, with age, SCAI-CSWG stage, LVEF, and lactate levels representing common denominators in risk assessment across both sexes.

In non-AMI-CS, convergent risk factors were: mechanical ventilation, SCAI-CSWG stages, platelets, and C-reactive protein. Nevertheless, women had more risk factors for the prediction of in-hospital mortality, such as age, BMI, and previous history of heart failure. While the CSWG registry reported higher BMI as associated with increased mortality in non-AMI-CS (21), our findings suggest a paradoxical protective effect of BMI in Latin American women. Also, our study is the first to highlight the prognostic relevance of liver function specifically in women, as albumin, ALT, and LDH emerged as independent predictors. Conversely, men showed a distinctive profile dominated by renal predictors: sodium, potassium, BUN, and diastolic blood pressure.

Treatment intensity and propensity scores matching insights

A notable sex gap has emerged in the intensity of treatment for AMI-CS. Across various SCAI stages, women consistently receive less aggressive treatment, such as PAC and hemodialysis, which are notably lower in women across SCAI E, contrasting with the CSWG experience (922). Given the established prognostic value of PAC in CS (222324), these disparities are unlikely to reflect biological differences and may represent treatment bias.

MCS use, particularly IABP, was also significantly higher in men in SCAI-CSWG D/E. This mirrors prior studies showing underutilization of MCS in women, often attributed to higher vascular complication risk and smaller vessel anatomy (69151625). Notably, our data provide additional insight into vasoactive selection: men were more frequently treated with norepinephrine and levosimendan, while women received vasopressin more often – a pattern not previously reported in AMI-CS.

In the non-AMI-CS, men exhibit a higher prevalence of PAC usage, especially in SCAI-CSWG-D, although MCS use was similar across stages, contrasting with the CSWG cohort, where men received more MCS in HF-CS (9). Women receive higher numbers of vasopressors (norepinephrine and vasopressin), whereas men were prescribed more inotropic (dobutamine and levosimendan); these differences, to our understanding, have not been revealed in previous studies (26).