Écrire un commentaire

General Impression

The preprint investigates the incremental prognostic value of atrial cardiomyopathy (AtCM) combined with cardiometabolic risk factor (CMRF) burden for predicting cardiovascular disease (CVD) mortality. The analysis utilizes the historical NHANES III cohort with a long follow-up period and includes baseline covariates such as lipids and creatinine.

Major Strengths:

• Cohort Scale and Follow-up Duration: The study utilizes a large population (NHANES III) with a follow-up period of nearly 14 years, providing substantial data for evaluating long-term cardiovascular mortality.

• Objective Baseline Data: The analysis incorporates standardized clinical examinations, laboratory assessments, and verified medication histories rather than relying solely on self-reported questionnaires.

• Controlled Covariates: The multivariable models include adjustment for baseline physiological factors, including lipid profiles and demographic variables.

Potential Impact:

• Risk Stratification: Evaluating atrial cardiomyopathy (AtCM) via accessible electrocardiographic (ECG) markers provides an additive approach to assessing long-term cardiovascular mortality risk.

• Clinical Awareness: The findings emphasize that electrical and structural changes in the atria carry independent prognostic significance in patients with an established metabolic burden. This underscores the potential utility of incorporating simple ECG screening into cardiovascular evaluations to better stratify risks among individuals with one or more cardiometabolic risk factors.

Major issues

1. Risk Evaluation within Subgroups

   The manuscript states that AtCM refines risk across specific strata based on the observation that AtCM was significantly associated with CVD mortality only among participants with risk factors, but not in the "No CMRF" subgroup (Page 9). However, the lack of statistical significance in the "No CMRF" group may be due to lower statistical power and fewer events in this subgroup, rather than a biological difference.

   Furthermore, the primary model demonstrates a non-significant multiplicative interaction term (P = 0.883), indicating that the prognostic association of AtCM is independent and additive rather than modified by the baseline cardiometabolic burden. Modifying the terminology (including the word "refines" in the title) to describe AtCM as an independent additive risk factor would align more closely with these statistical results.

2. Assessment of Renal Function

   The multivariable models utilize raw serum creatinine instead of the estimated glomerular filtration rate (eGFR) as a renal covariate. As shown in Table 1, female participants predominate in specific strata (e.g., 66.4% in the ≥2 CMRFs / AtCM- group). Because serum creatinine levels vary by age, sex, and muscle mass, crude values can mask underlying kidney dysfunction in older or female cohorts.

   Computing eGFR using a standard formula (such as CKD-EPI) and including it as a continuous covariate would provide more precise control for renal function.

Minor issues

1. Classification of Smoking Status

   Smoking is adjusted for as a binary variable ("current smoking": yes/no). Former smokers with a history of cumulative exposure are grouped together with individuals who have never smoked. Addressing this potential source of residual confounding in the limitations section would clarify the model assumptions.

2. Methodological Implications of the Long Follow-Up Period

   Given the nearly 14-year follow-up, two clinical factors affect the long-term risk assessment. First, a proportion of the cohort died from non-cardiovascular causes (such as cancer), which affects the calculation of long-term cardiovascular risk. Second, participants' metabolic risk factors and ECG parameters may have changed over time, introducing potential misclassification bias. Expanding the limitations section to discuss how non-cardiovascular mortality and over-time clinical changes might affect the observed Hazard Ratios would provide relevant context.

3. Generalizability and Evolution of Hypertension Management

   The Discussion notes that findings from the historical NHANES III cohort (1988–1994) may not fully generalize to contemporary populations due to temporal changes in cardiometabolic disease management. Specifying the pharmacological nature of this shift would clarify this point. While the historical data reflect an era reliant on diuretics and beta-blockers, modern clinical practice utilizes agents—such as renin-angiotensin-aldosterone system (RAAS) inhibitors and angiotensin receptor-neprilysin inhibitors (ARNIs)—that directly mitigate myocardial tissue remodeling and reverse left atrial fibrosis. Mentioning these specific therapeutic shifts would clarify the relevance of the observed AtCM prognostic risk for modern practice.

4. Missing Participant Flowchart

   The manuscript does not include a participant flowchart detailing the sample attrition at each step of the exclusion process. Providing a flowchart stating the exact numbers of participants excluded due to missing variables, absence of sinus rhythm, baseline CVD, or BMI < 18.5 kg/m² would ensure complete transparency regarding the final sample size (n = 7,083).

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they did not use generative AI to come up with new ideas for their review.