Non‐alcoholic fatty liver disease and the risk of dementia

With great interest, we read the recent study by Kim and coworkers showing that nonalcoholic fatty liver disease (NAFLD) is associated with an increased risk of dementia in a nationwide cohort study.1 We thank them for their significant work for clinical practice, but there exist some potential pitfalls which may affect the credibility of the results. Firstly, the baseline characteristics between NAFLD and nonNAFLD group were severely unbalanced, although multivariable analysis and propensity score matching analysis was performed to reduce confounders. However, the difference in the risk of dementia may originate from some other uncontrolled confounders, such as smoking, BMI, diabetes, drugs (including anticholinergic and sedativehypnotic medications), psychosocial and traumarelated stress (including previous traumatic brain injury, loneliness and social isolation). Especially, previous studies showed that sedentary lifestyle was a significant predictor for NAFLD.2 Also, accumulating evidence revealed that sedentary lifestyle was associated with an increased risk of dementia.3 Thus, these important risk factors including sedentary lifestyle should be taken into account in multivariable analysis and propensity score matching analysis. Secondly, some evidence tends to support the notion that NAFLD does not increase the risk of dementia. Shang and colleagues found that NAFLD was not associated with the risk of dementia.4 Labenz and coworkers showed that there was no association between NAFLD and the incidence of allcause dementia on Cox regression analysis.5 Based on the current conflicting evidence, we performed a metaanalysis to explore the association between NAFLD and risk of dementia. Adjusted hazard ratios and 95% CI were pooled to calculate an overall estimate of effect size. All statistical analyses were performed with Stata 12.0 (Stata Corp). A twotailed p value <.05 was considered significant. A total of four observational studies were included in the metaanalysis through systematic literature search from PubMed, Embase and Web of science until 20 April 2022.1,4– 6 Pooled analysis showed that NAFLD was not linked with a higher risk of dementia on Random effect model (HR, 1.02; 95% CI, 0.91– 1.14; Figure 1A), which was consistent with the results of the ‘leaveoneout’ sensitivity analysis (Figure 1B). In summary, we respect the great work done by the authors, but the study should be interpreted with the aforementioned limitations. Further, highquality studies should be warranted to clarify the correlation between NAFLD and the risk of dementia before we use the findings in clinical practice. FUNDING INFORMATION None.