We read with interest the article by Zhang et al. [1] describing gut microbial and metabolomic alterations in cirrhotic portal hypertension (CPH). While the integrative design is commendable, several methodological concerns limit the strength of the conclusions.
First, the cohort was small (35 CPH vs. 71 controls) and aetiology-homogeneous, with all patients from a single hepatitis B–dominant center. Microbiota profiles differ across cirrhosis etiologies and geographies, limiting external validity.
Second, the choice of controls—non-cirrhotic chronic hepatitis B—introduces potential confounding. Many had active hepatic inflammation, which independently shapes microbiota and metabolome. Subgroup analysis in virologically suppressed controls reduced power and led to the loss of significance in numerous markers. Prior work shows that even viral eradication does not normalize dysbiosis in cirrhosis [2].
Third, unmeasured confounders weaken causal inference. Diet, medications such as proton pump inhibitors, and comorbidities strongly influence microbial composition. Chronic opioid use, for instance, predicts microbial shifts and worse outcomes in cirrhosis [3]. The absence of such adjustments makes attribution to portal hypertension uncertain.
Fourth, the murine model offers only partial translational relevance. N-nitrosamine–induced disease does not reproduce human hemodynamics, and culture-based isolation of Veillonella and Streptococcus risks selection bias. Clinical experience with probiotics illustrates that microbiome modulation alters metabolomic pathways but does not automatically translate into durable outcome gains [4].
Finally, multiplicity remains a concern. Testing over 120 genera and 40 metabolites risks false positives, even with adjustment. Without an independent validation cohort, biomarker robustness remains unproven.
In summary, Zhang et al. provide valuable exploratory data on gut–liver interactions in portal hypertension. However, small sample size, limited external validity, confounded control selection, and lack of validation necessitate cautious interpretation. Multicenter, aetiology-diverse cohorts with dietary and pharmacological adjustment and replication are essential before therapeutic recommendations can be justified.
Muhammad Talha Ghaffar: conceptualization, writing – original draft, writing – review and editing, visualization.
The author has nothing to report.
The author has nothing to report.
The author declares no conflicts of interest.
This article is linked to Zhang et al. papers. To view these articles, visit https://doi.org/10.1111/apt.70392.
Data sharing not applicable to this article as no datasets were generated or analysed during the current study.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：