Letter on “Autoimmune Hepatitis and Vitamin D Deficiency: A Nationwide Perspective”

Editors,

We read the paper “Autoimmune Hepatitis and Vitamin D Deficiency: A Nationwide Perspective”, published in Alimentary Pharmacology and Therapeutics. I highly appreciate the author Kilani's efforts in discovering the relation of vitamin D deficiency with disease outcomes in patients with autoimmune hepatitis [1]. However, after careful evaluation of this study, we have some concerns related to the methodology and clinical aspects, which can affect the outcomes of this study and persuaded us to write this letter.

To compare different AIH patient groups based on their vitamin D levels, the authors ran several different statistical tests, such as propensity score matching (PSM), comparative analysis, significance, and statistical tests for baseline without using any adjustment tool, leading to increased false positives and noise levels in the study [2]. Further, using PSM for the sake of controlling confounding variables and selection bias can cause a significant loss of records, as exemplified in this study, reducing sample size and falsifying the actual relation of vitamin D levels with poor outcomes in AIH patients [3]. Further, confounders like sunlight exposure, physical activity, and diet, which can significantly influence vitamin D levels, were not accounted for. This negligence can introduce residual confounding because the actual deficiency of vitamin D can be misinterpreted as the deficiency caused by these factors, thereby raising questions about the reliability of PSM and the outcomes of this study [4, 5]. The more severe the disease, the poorer its outcomes. However, the study did not cover the severity of AIH in patients, which greatly impacts clinical outcomes independent of vitamin D deficiency. Patients with severe AIH were reported to have higher mortality rates and need for liver transplantation [6].

Furthermore, immunosuppressive drugs, such as corticosteroids and azathioprine, which are used in standard care for the treatment of AIH, can worsen disease outcomes regardless of vitamin D levels. Studies show that azathioprine and corticosteroids are both associated with adverse consequences, including cataracts, diabetes, hypertension, gastrointestinal disturbance and increased hospitalisation in AIH patients [7, 8]. It might be possible that the adverse outcomes associated with AIH are caused by these therapies instead of vitamin D deficiency.

In conclusion, by using multivariate models with propensity score matching, future studies can address limitations of propensity score matching, such as residual confounding and the inability to retain non-matched individuals. Using Bonferroni Correction as an adjustment can address false positives that arise by running multiple statistical analyses. Lastly, future studies should analyse AIH severity, measure the impact of factors like sunlight and physical activity, and provide the limitations associated with immunosuppressants.

Syed Tawassul Hassan: conceptualization, writing – original draft, validation, visualization, resources. Muhammad Shaheer Bin Faheem: supervision, project administration, writing – review and editing, validation, conceptualization, writing – original draft. Muhammad Rehan Zahid: resources, project administration.

The authors have nothing to report.

The authors declare no conflicts of interest.

This article is linked to Kilani et al paper. To view this article, visit https://doi.org/10.1111/apt.18438.