Clinical Phenotyping and Treatment - Related Considerations in the Prognostic Assessment of the BNP - to - Albumin Ratio [Letter]

We read with interest the article by Zeng et al entitled "The BNP-to-Albumin Ratio (BAR) as a Predictor of All-Cause Mortality in Elderly Patients with Atrial Fibrillation." 1 The study examines an inexpensive marker that combines cardiac hemodynamic stress with nutritional and inflammatory status.The reported association between elevated BAR and allcause mortality supports further investigation, but several clinical and methodological issues may affect the interpretation of its independent prognostic value.The high-BAR group appears to represent a substantially more advanced heart failure phenotype.Heart failure was present in 61.5% of patients in the high-BAR group and 23.4% in the low-BAR group.The high-BAR group also had a lower median left ventricular ejection fraction, 55% (interquartile range, 46%-66%) versus 65% (56%-71%), larger cardiac chamber dimensions, lower estimated glomerular filtration rate, and higher high-sensitivity C-reactive protein and troponin concentrations.Although heart failure was included as a binary variable in the multivariable models, the authors did not report the heart failure phenotype according to LVEF category.The group-level median LVEF cannot classify the individual patients with heart failure, particularly because the interquartile range indicates considerable heterogeneity.Treatment information is also needed to interpret this imbalance.The Methods section states that data on diuretics and sodium-glucose cotransporter-2 inhibitors were collected, but their use was not presented in the main results or Table 1.Mineralocorticoid receptor antagonist therapy was not reported, and angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and angiotensin receptor-neprilysin inhibitors were combined into one category.The proportion receiving sacubitril-valsartan therefore cannot be determined.Heart failure phenotype and evidence-based therapy may affect both BNP concentrations and long-term mortality.Reporting the distribution of heart failure phenotypes and the use of SGLT2 inhibitors, mineralocorticoid receptor antagonists, diuretics, and sacubitril-valsartan across BAR groups would be valuable.A sensitivity analysis incorporating these variables would strengthen the findings.A second clinically relevant imbalance concerns atrial fibrillation status and management.Patients in the high-BAR group had a higher heart rate than those in the low-BAR group, with means of 94.45 (SD 23.66) and 83.75 (SD 19.13) beats per minute, respectively.They also had a markedly lower prevalence of AF ablation, 9.2% versus 24.6%, despite more frequent beta-blocker use, 78.0% versus 60.9%.Baseline heart rate was not incorporated into the fully adjusted mortality model.The authors also did not report AF pattern, rhythm at the time of BNP measurement, adequacy of ventricular rate control, or the use of other rate-control therapies such as digoxin.BNP concentrations may be influenced by AF burden, rhythm status, ventricular rate, and associated hemodynamic stress.2 Elevated BAR may therefore partly identify patients with uncontrolled or more advanced AF.The lower ablation rate should not be interpreted causally because it may reflect differences in age, comorbidity burden, AF substrate, procedural eligibility, or patient selection.Clarification of whether ablation occurred before or during the index hospitalization, together with the distribution of paroxysmal, persistent, and permanent AF, would improve the clinical