The study protocol adhered to the Declaration of Helsinki and was approved by the institutional review board of the Yonsei University Health System. Written informed consent was obtained from all patients included in the Yonsei AF Ablation Cohort Database (NCT02138695). Among 3,375 patients who underwent de novo AFCA for symptomatic drug-refractory AF, 3,239 patients who had available data on their serum albumin, cholesterol, and total lymphocyte count at the time of the ablation procedure were enrolled in this study (cohort 1). All patients stopped their AADs for a period of at least five half-lives before the catheter ablation. The exclusion criteria were as follows: (1) permanent AF refractory to electrical cardioversion, (2) AF with valvular disease ≥grade 2, and (3) a previous cardiac surgery with a concomitant AF surgery or AFCA.

For external validation of the association between nutritional status and outcomes of catheter ablation, we used an independent AF ablation cohort that included 805 patients who underwent their first AFCA at Korea University Cardiovascular Center from 2013 to 2016. From the independent data, we enrolled and analyzed 360 patients who had available data on their serum albumin, cholesterol, and total lymphocyte [median age 57 (interquartile range 50-64) years, 20.8% female, and 46.4% paroxysmal AF] for CONUT score evaluation (cohort 2).

A diagnosis of malnutrition was reached using the CONUT score, which is a screening tool for the nutritional status of hospitalized patients (18) and was calculated according to the levels of serum albumin, cholesterol, and total lymphocyte count (Table 1). High CONUT score has been known to have a prognostic impact in patients with chronic cardiac disease including heart failure and AF (6, 14). Also, the score has been associated with outcomes after cardiovascular surgery (19) or interventional procedures including transcatheter aortic valve replacement and percutaneous coronary intervention (20, 21). A score of 0–1 was considered normal (good) nutrition, whereas malnutrition could be classified as mild (2–4) or moderate to severe (≥5).

Intracardiac electrograms were recorded using the Prucka CardioLab Electrophysiology system (General Electric Medical Systems, Inc., Milwaukee, WI, USA). Three-dimensional electroanatomic mapping (NavX, St. Jude Medical, Inc., Minnetonka, Minnesota and CARTO, Biosense-Webster, Inc., Diamond Bar, California) was performed using a circumferential pulmonary vein (PV) mapping catheter (Lasso, Biosense-Webster Inc.) through a long sheath (Schwartz left 1, St. Jude Medical, Inc.). Transseptal punctures were performed, and multi-view pulmonary venograms were obtained. The details of the AFCA technique have been described in our previous studies (22, 23). All patients underwent a circumferential PV isolation (CPVI) during the de novo procedure. Most patients (87.8%) underwent the creation of cavotricuspid isthmus block during the de novo procedure, unless there was atrioventricular conduction disease. As an extra-PV left atrial (LA) ablation, additional linear ablation, including a roof line, posterior inferior line (posterior box lesion), and anterior line, was performed, especially in patients with persistent AF. A left lateral isthmus ablation, right atrial ablation, or complex fractionated electrogram ablation were performed in a minority of patients at the operator's discretion. An open-irrigated tip catheter [Celsius (Johnson & Johnson, Inc., Diamond Bar, CA); NaviStar ThermoCool (Biosense Webster, Inc); ThermoCool SF (Biosense Webster, Inc); ThermoCool SmartTouch (Biosense Webster, Inc); Coolflex (St. Jude Medical, Inc); 30–35 W; 47°C; FlexAbility (St. Jude Medical, Inc); ThermoCool SmartTouch (Biosense Webster, Inc.); and TactiCath (St. Jude Medical, Inc)] was used. Systemic anticoagulation was attained with intravenous heparin to maintain an activated clotting time of 350–400 s during the procedure.

After completion of the protocol-based ablation, the procedure was completed when no recurrence of AF was observed within 10 min after cardioversion with an isoproterenol infusion (5–10 μg/min depending on ß-blocker use, target sinus heart rate of 120 bpm). If further AF triggered or frequent unifocal atrial premature beats were observed due to the effect of isoproterenol, extra-PV foci were ablated using quick 3D activation mapping. We defined an extra-PV LA ablation as an additional linear ablation with or without a complex fractionated electrogram ablation following the CPVI. If recurrent AF or atrial tachycardia (AT) were seen repeatedly under AADs after the de novo AFCA, we recommended a repeat AFCA. The detailed technique and strategy for repeat ablation procedures were presented in a previous study (22, 23).

We discharged patients not taking AADs except for those who had recurrent extra-PV triggers after the AFCA procedure, symptomatic frequent atrial premature beats, non-sustained atrial tachycardia, or an early recurrence of AF on telemetry during the admission period (28.7%). Electrocardiography was performed for all patients visiting an outpatient clinic 1, 3, 6, and 12 months after AFCA and every 6 months thereafter or whenever symptoms developed. Twenty-four-hour Holter recordings were performed at 3, 6, and 12 months and every 6 months thereafter. Patients who reported episodes of palpitations suggestive of arrhythmia recurrence underwent Holter monitoring or event monitor recordings. AF recurrence was defined as any episode of AF or AT lasting for at least 30 s. Any ECG documentation of AF recurrence within a 3-month blanking period was diagnosed as an early recurrence, and AF recurrence occurring more than 3 months after the procedure was diagnosed as clinical recurrence.

Continuous variables are summarized as medians (interquartile range) and were compared using the Mann-Whitney U-test or Kruskal-Wallis H-test. Categorical variables are summarized as frequencies (percentages) and were compared using Fisher's exact test. We used Cochran-Armitage analysis to investigate trends in the complications of AFCA according to CONUT score. Multivariate logistic regression was applied to identify predictors associated with the occurrence of overall and major complications after AFCA. Kaplan–Meier analysis with the log-rank test was used to calculate AF recurrence-free survival over time and to compare recurrence rates across groups. A two-sided P-value of < 0.05 was considered statistically significant. The statistical analyses were performed using R version 4.0.2 software (The R Foundation, www.R-project.org).

Table 2 summarizes the baseline clinical and procedure-related characteristics of the patients undergoing AFCA according to their nutritional status. Of the 3,239 patients included [median age 59 (interquartile range 52-66) years, 26.8% female, and 67.7% paroxysmal AF] (cohort 1), 1,005 (31.0%) had malnutrition; 991 (30.6%) had mild malnutrition (CONUT scores: 2–4), and 14 (0.4%) had moderate-to-severe (CONUT scores: ≥5) malnutrition. Patients with malnutrition were more likely to be older (P < 0.001) and male (P = 0.014), and had a lower BMI (P < 0.001), lower blood pressure (P < 0.001), higher CHA₂DS₂-VASc score (P < 0.001), and more comorbidities than those with normal nutritional status.

The rates of complications of AFCA according to CONUT score in cohort 1 are summarized in Figure 1. The overall complication (P for trend = 0.037) and major complication rates (P for trend = 0.028) were greater in individuals with higher CONUT scores. Detailed information about the complications of AFCA according to nutritional status are presented in Table 3. During the first half of the study period (2009–2015), the overall complication rates after AFCA were 4.0 and 5.1% in those with normal nutrition and malnutrition (CONUT scores: ≥2), respectively. During the second half (2015–2020), the rates slightly decreased to 3.6% for a normal nutrition and 4.9% for a malnutrition. The overall complication rates after AFCA were 3.3% for a normal nutrition, 4.2% for a mild malnutrition, and 21.4% for a moderate-to-severe malnutrition. The major complication rates were 1.9, 2.6, and 14.3% for normal nutrition, mild malnutrition, and moderate-to-severe malnutrition, respectively. In multivariable logistic regression, moderate-to-severe malnutrition was associated with an increased risk of overall (OR 6.456, 95% CI 1.637–25.463, P = 0.008) and major complications (OR 5.845, 95% CI 1.164–29.359, P = 0.032), compared to normal nutritional status (Table 4). Older age (OR 1.020 per 1-year increase, 95% CI 1.001–1.039), female sex (OR 1.915, 95% CI (1.302–2.817), and higher systolic blood pressure (OR 1.013 per 1-mmHg increase, 95% CI 1.000–1.026) were also independent predictors for the occurrence of complications after AFCA.

Figure 2 shows the effects of nutritional status on incidences of AFCA complications according to body mass index and sex. Even among overweight or obese patients with body mass indices at least 25 kg/m², malnutrition was prevalent (27.8%). Regardless of body mass index, malnutrition showed trend toward a higher risk of complications compared with normal nutrition (Figure 2A). The rates of overall complications after AFCA were highest (7.1%) in malnourished female patients, followed by 5.3% in normally nourished females, 3.7% in malnourished male patients, and 2.5% in normally nourished males (P for trend < 0.001) (Figure 2B).

Of the 360 patients in the validation cohort (cohort 2), 120 (33.3%) had malnutrition; 117 (32.5%) had mild and 3 (0.8%) had moderate-to-severe malnutrition. 26 (7.2%) had complications after AFCA. The baseline clinical and procedure-related characteristics of the patients in the validation cohort are presented in Supplementary Table 1. The overall complication rates after AFCA were 5.0 and 11.7% for normal nutrition and malnutrition, respectively. Detailed information about the complications of AFCA are presented in Supplementary Table 2. After multivariable adjustment, malnutrition (CONUT ≥ 2) was an independent predictor (OR 2.874, 95% CI 1.174-7.033) for complications after AFCA (Supplementary Table 3). In addition, an increasing CONUT score was associated with a higher risk for complications (OR 1.418 per 1 point increase, 95% CI 1.049-1.916).

Among 3,239 patients in the main cohort (cohort 1), neither the early recurrence rate within 3 months of the AFCA (32.6 vs. 34.4%, P = 0.331) nor the clinical recurrence rate (37.0 vs. 36.6%, P = 0.829) differed between the normal nutrition and malnutrition groups during the median follow-up of 40 months (interquartile range 18–74 months, Table 5). Kaplan–Meier analysis showed no significant difference in overall AF recurrence (log-rank P = 0.763; Figure 3A) or AAD-free AF recurrence (log-rank P = 0.148; Figure 3B) between the groups. Repeat ablation procedures were performed in 13.3% of the normally nourished patients and 11.7% of the malnourished patients (P = 0.230). In those with a repeat ablation, the proportion of PV reconnections did not differ according to nutritional status (P = 0.396, Table 5).

In the validation cohort of 360 patients (cohort 2), there were consistently no differences in overall AF recurrence or AAD-free AF recurrence between the normal nutrition and malnutrition groups during the median follow-up of 14 months (interquartile range 6-28 months, Figures 3C,D).

Malnutrition, defined by CONUT score, is relatively common in patients with AF undergoing catheter ablation. In this study, we noted a trend toward a higher risk of AFCA complications in patients with higher CONUT scores. Patients with moderate-to-severe malnutrition (CONUT scores ≥ 5) and females with CONUT scores ≥2 faced a substantially higher risk of complications after AFCA. In an external validation cohort, malnutrition was consistently associated with an increased risk of complications. There was no significant association between nutritional status and AF recurrence after AFCA.

Using the same scoring as that in the present study (CONUT score), Zhu et al. reported that 36.6% of patients undergoing AFCA were malnourished, and 2% had moderate-to-severe malnutrition (15). The results of that study are consistent with those of the present study: 31% of the patients undergoing AFCA had malnutrition, although only 0.4% had moderate-to-severe malnutrition. Several studies have investigated the association of BMI with the clinical outcomes in patients with AF, in which they reported that a BMI of at least 25 kg/m² was associated with a lower risk of a stroke, cardiovascular death, or all-cause mortality, suggesting an “obesity paradox” in AF (5). However, research evaluating the prognostic value of nutritional status beyond BMI is scarce. Raposeiras-Roubín et al. reported that malnutrition, defined using CONUT score, is more prevalent (43.1%) in octogenarian patients with AF and that malnutrition is associated with a higher risk of mortality, strokes, and major bleeding (14). In this study, malnutrition was prevalent even in those with body mass indices at least 25 kg/m². Those findings suggest a discrepancy between BMI and nutritional status and support the importance of assessing and improving nutritional status beyond BMI. The negative impact of malnutrition on the course of AF might be attributable to cardiac cachexia, which leads to the activation of proinflammatory cytokines and neurohormonal dysfunction (24) and a weakened protective effect of adipokines stemming from a reduced fat mass (25).

Numerous studies have investigated relationships between nutritional status and AFCA outcomes using BMI to evaluate nutritional status, and all of them have focused on the efficacy of AFCA, that is, the AF recurrence rate. Baek et al. reported that being overweight and having metabolic syndrome were associated with a higher long-term recurrence rate at 2 years after AFCA in Asian populations in which the proportion of patients with a BMI > 30 kg/m² was only 5.5% (16). Further, Deng et al. reported that being underweight (BMI < 18.5 kg/m²) was also associated with a higher AF recurrence (4). However, Bunch et al. reported that there was no significant association between BMI and AF recurrence (26). A recent study in 246 patients undergoing AFCA on the prognostic value of malnutrition using two nutritional status screening tools, including CONUT score, reported that malnutrition was an independent predictor of the recurrence of AF (15). In the present study, we expounded on the prior observations by enrolling a substantially larger number of participants and using longer follow-up durations, showing that the rates of early (<3 months) and clinical (beyond 3 months) AF recurrence did not differ according to nutritional status. However, the complication rates increased significantly as CONUT score increased. Up to 2 out of 10 patients with a CONUT score ≥ 5 sustained complications after AFCA. The trend was consistently observed in overweight or obese patients with AF. Female sex has been reported to be a risk factor for AFCA complications (27), which was consistently shown in this study. The present study showed a substantial risk for AFCA complication in malnourished females. Thus, emphasis should be placed on the assessment of nutritional status in patients with AF who are scheduled to undergo catheter ablation.

This study had several limitations. First of all, the proportion of those with moderate-to-severe malnutrition was too small (0.4%) in this study. Although they were at significantly higher risk of complications in multivariable analyses, this needs to be interpreted carefully and cannot be generalized. This was a retrospective observational cohort study from two centers that included patients using strict selection criteria referred for AF ablation; hence, our findings cannot be used to establish causal relationships. Also, we were not able to fully exclude selection bias. Nutritional status was measured using CONUT score instead of studying body composition or detailed information on diet. Some patients whose CONUT score elements were not available were excluded from this study. We did not compare the prognostic values of other validated screening tools for malnutrition.