With the aging of the worldwide population and advances in medical care, physicians face an ever-growing number of elderly patients with advanced atherosclerotic lower extremity arterial disease (LEAD) (1, 2). Patients with LEAD have a 3- to 4-fold risk of incident cardiovascular (CV) events, even asymptomatic patients (3). Symptomatic patients have a higher risk of subsequent CV events, ranging from 20% for 5-year non-lethal CV events in claudication to 10% for 1-year deadly events in patients with chronic limb-threatening ischemia (CLTI) (4, 5), especially in elderly patients with multiple comorbidities (6–8).

Corrected QT interval (QTc) varies depending on gender and age, with longer QTc intervals in older people and women (9). Population-based studies have shown the association between CV events and death with longer QTc intervals in healthy populations (10–12) and patients with increased CV risk (13, 14). For atherosclerotic LEAD, limited reports have shown the impact of longer QTc intervals on mortality or non-fatal CV events in patients with diabetic foot ulcers (15, 16). The role of longer QTc intervals in elderly patients with symptomatic LEAD has not yet been elucidated. This study aimed to examine the impact of QTc intervals on long-term CV outcomes in older people with symptomatic atherosclerotic LEAD.

We presented continuous data with or without a normal distribution as mean ± standard deviation or median and interquartile range, which were compared using one-way ANOVA or the Kruskal–Wallis test. Categorical data were shown as frequencies and percentages and compared using the chi-squared test. We determined the cutoff value of continuous variables for main outcomes using the receiving-operative characteristic curve. We used the Kaplan–Meier method for time-to-event analyses with censoring. We examined the impact of the QTc interval on the main outcome using a multivariate Cox proportional hazards model. The covariates in the multivariate models contained demographic factors (age, sex, current smoking), traditional CV risk factors [body mass index (BMI), dialysis, diabetes mellitus, and hypertension], previous cardiovascular disease history (CAD, HFrEF, CVA, CLTI, and atrial fibrillation), actual laboratory results [white blood cell count, neutrophil-lymphocyte ratio, C-reactive protein (CRP), and serum albumin], and medication (antiplatelet agents, β-blockade, renin-angiotensin-aldosterone system inhibitors, and statins). We included one variable per 10 observed events to prevent model overfitting. We also analyzed the interaction between QTc interval and other significant predictors of all-cause mortality and MACEs. The software used for statistical analyses were Statistical Package for the Social Sciences, version 22.0 (SPSS Inc., Chicago, IL, USA), R software (version 3.4.3; The R Foundation for Statistical Computing, Vienna, Austria), and MedCalc® Statistical Software, version 20.110 (MedCalc Software Ltd, Ostend, Belgium). A P < 0.05 was considered as a statistical significance.

A total of 504 patients were entered into data analysis [235 men (46.6%); mean age, 79.9 ± 6.2 years; mean QTc interval, 459 ± 33 msec]. Patients were categorized according to terciles of QTc intervals (Tercile 1: 353–442 msec; Tercile 2: 443–471 msec; Tercile 3: 472–568 msec). Table 1 summarizes the baseline characteristics of the tercile groups. Compared with the other groups, patients in tercile 3 had a higher proportion of low BMI, advanced limb ischemia, and medical comorbidities, such as CAD, HFrEF, and dialysis dependence. We also found that the tercile 3 group had less use of RAASi medication, a lower serum albumin and potassium level, and a higher level of immune-inflammatory markers (CRP, neutrophil-lymphocyte ratio, and lymphocyte count).

The median follow-up period was 3.15 (IQR, 1.65–5.42) years, and we noted 264 deaths (183 non-cardiac deaths and 81 cardiac deaths) and 145 MACEs. Most older people die from infectious complications and multi-organ dysfunction, followed by fatal CV events. Eighty-three patients have subsequent coronary interventions during the follow-up period, including 56 acute coronary syndromes and 27 angina pectoris. The 5-year cumulative incidences of following coronary interventions were significantly different among the three tercile groups (12.6%, 17.6%, and 36%, P < 0.001) (Supplementary Figure). Kaplan-Meier analysis showed that the 5-year rates of freedom from all-cause mortality (71% vs. 57% vs. 31%, P < 0.001) and MACEs (83% vs. 67% vs. 46%, P < 0.001) were significantly different among the three tercile groups (Figures 2A,B). Table 2 shows the relationship between QTc interval and clinical outcomes (all-cause mortality and MACE). In univariate analysis, a 1-SD increase in QTc interval increased the unadjusted risk of all-cause mortality (hazard ratio [HR] 1.74, 95% confidence interval [CI] 1.54–1.98, P < 0.001) and MACEs (HR 1.20, 95% CI 1.14–1.26, P < 0.001) at 5 years. Based on the categorical variables, the risk of all-cause mortality and MACEs increased with a prolonged QTc interval. Compared with tercile 1, the HRs in tercile 3 were 3.81 for death and 4.84 for MACEs (P < 0.001). Even after adjusting for clinical and laboratory covariates, the influence of QTc interval on 5-year all-cause mortality and MACEs remained statistically significant.

Table 3 shows the multivariate analysis of the outcome of interest. Age, BMI, CAD, CVA, QTc interval, CRP, and albumin were independent factors for all-cause mortality, and predictors for MACEs were HFrEF, CAD, CVA, QTc interval, and CRP. We performed an interaction analysis between QTc intervals and other significant covariates. The results showed that the QTc interval remained a significant predictor for primary outcomes and that the QTc interval and CRP level were most strongly associated with death (HR = 4.89, 95% CI 3.09–7.73, interaction P < 0.001) and MACEs (HR = 7.83, 95% CI 4.14–14.79, interaction P < 0.001). The scatter plot (Figures 3A,B) reveals that longer QTc intervals can independently predict the outcome and synergistically affect death and MACEs when combined with other predictors. For example, we found the highest rate of all-cause mortality (77%) in patients with QTc >465 msec combined with CRP level >2.1 mg/dl and MACEs (39%) in patients with QTc >463 msec coupled with CRP >0.88 mg/dl.

This cohort study is the first to demonstrate the association between longer QTc intervals and a rising risk of incident all-cause deaths and MACEs in older people with symptomatic LEAD. Patients with longer QTc intervals had advanced lower limb ischemia, multiple comorbidities, and impaired nutritional and immune-inflammatory status. After multivariate analysis, we noticed that a longer QTc interval was not only an independent predictor but also had a synergistic predictive power on the primary outcome when combined with other covariates.

The Framingham Heart Study showed a graded correlation between QTc interval and all-cause death, CAD-related death, and sudden cardiac death (17). Previous research has also demonstrated a similar association between QTc interval and stroke or peripheral vascular disease (10, 18). Patients with symptomatic LEAD, from disabling claudication to CLTI, carry a higher risk of death and CV events (4, 5), especially elderly patients due to multiple comorbidities, frailty, and poor candidacy for revascularization. Previous studies have shown that older age, CAD, atrial fibrillation, dialysis, CLTI, stroke, malnutrition, non-ambulance use, and a higher neutrophil-lymphocyte ratio are potential predictors of subsequent death and cardiovascular events in elderly patients after revascularization for LEAD (19–21). Kula et al. also reported a strong association of high-sensitivity troponin T and N-terminal-pro-B-type natriuretic peptide with QTc prolongation in patients with chronic kidney disease (CKD). They proposed the crucial role of myocardial ischemia and volume overload in conduction delay and future MACEs in those patients (22). Compared to the tercile 1 group, our tercile 2 and 3 groups have a higher proportion of CKD, dialysis dependence, HFrEF, and CAD, which implied the association of QTc prolongation with myocardial stretch and ischemia. And thus, the tercile 1 group has better survival and fewer MACEs.

Data regarding the association between QTc interval and symptomatic LEAD are scarce. Fragher et al. reported that QTc interval prolongation (>440 ms) increased mortality in 70 diabetic patients who underwent above-the-ankle amputation. The 3-year all-cause mortality rate was 54% in their study (23). Another study also revealed that a longer QTc interval (>440 msec) combined with a lower HbA1c level (<7.5%) could forecast a higher all-cause mortality in 240 patients with diabetic foot ulcers (15). Wang et al. collected data from 331 patients with foot ulcers with an average 4-year follow-up. They found that QTc prolongation independently predicted cardiac death (HR 5.465, P = 0.039) but not ulcer healing or recurrence (16).

However, the role of QTc in elderly patients has remained unclear. In this study, we disclosed that a longer QTc interval was strongly associated with all-cause death and MACEs; either QTc interval was used as a continuous variable or tercile categorization. The underlying mechanism between QTc prolongation and CV events has yet to be elucidated. Cardiac autonomic neuropathy (CAN), a severe complication of diabetes, strongly correlates with a roughly five-fold risk of CV mortality (24). Previous research has reported that CAN could be proposed as an imbalance between sympathetic and parasympathetic activity in the myocardium and lower extremities (10, 25, 26). QTc interval prolongation has been used for CAN assessment and might correlate with autonomic failure in diabetic patients (24, 27). In our study, the tercile 3 group had the most prolonged QTc interval and higher proportions of type 2 DM, CAD, HFrEF, dialysis, advanced limb ischemia, malnutrition, elevated NLR, and CRP. These baseline characteristics are common risk factors for CAN. A longer QT interval might also reflect a reduced repolarization reserve, possibly due to sicker elderly patients with many medical comorbidities. Higher sympathetic tone and profound CAN could accelerate atherosclerosis and further predispose these patients to death and CV events.

Multivariate analysis showed that the QTc interval is independent in predicting death and MACEs. Moreover, interaction analysis and scatter plots revealed a synergistic effect between QTc and other covariates for outcome prediction. QTc interval and CRP level were most strongly associated with death and MACEs. Inflammation is crucial for initiating and developing atherosclerosis and triggers CV events (28). CRP is the widely used inflammatory biomarker linked to underlying atherosclerosis and future CV disease. The association between CRP levels and QTc prolongation has been reported in patients with systemic inflammation or hypertension (29–31). Inflammatory biomarkers, such as interleukin-6, prolong the ventricular myocyte action potential by inhibiting the hERG potassium channel in cardiac myocytes, proposing a direct mechanistic link between QTc prolongation and inflammation (32). Symptomatic LEAD included advanced atherosclerosis and tissue inflammation, which might explain the essential role of combined CRP and prolonged QTc interval in predicting death and MACEs.

In elderly patients with symptomatic atherosclerotic LEAD, a prolonged QTc interval is associated with advanced LEAD, multiple medical comorbidities, impaired nutrition, and immune-inflammatory status. QTc interval prolongation increases the risk of future MACEs and all-cause mortality and may be considered in the risk assessment of these patients.