In our analysis of 100 subjects, we observed a negative correlation between αS and steatosis, but also a significant negative correlation between liver fibrosis and αS in the brain. When assessing correlations between the extent of pathologies, a significant negative correlation was noted between the extent of LP, liver steatosis, or fibrosis, and the extent of αS in the brain. Thus, both categorical data and the extent of pathologies revealed a negative correlation between LP and αS pathology. Furthermore, there was a significant negative association between the observation of Alzheimer type II astrocytes and αS pathology in the brain. Based on our results, i.e., a negative correlation between LP and αS in the brain, it seems that subjects with LP are not predisposed to develop PD, PD /DLB; rather, the opposite. This outcome is in line with the recent large meta-analysis reporting an inverse association between alcohol consumption and PD [22, 37]. The etiopathogenesis of this outcome is unclear. Gastrointestinal tract is altered in subjects with HAC and it has been reported that moderate acute alcohol consumption immediately damages the enterocytes [12, 15, 45]. In parallel αS is observed in the gastrointestinal nervous system years prior to be seen in the brain [25]. Thus, the question does arise whether eventual alcohol-related alterations in the gastrointestinal tract influence the development of αS related alterations seen in the neuronal cell population of the gut. Moreover, it is well known as was also seen by us that in subjects with HAC the glial cells are altered in the brain but nothing is known regarding the glial cell population in the gut. Furthermore, it is not clear how severe LP might change livers role in the suggested clearance of αS [47].

In our analysis, we found no association between HPτ and any of the assessed liver pathologies. However, there was a significant negative correlation between liver steatosis and Aβ. Here, we did not identify any association between various extents of LP and the extent of Aβ in the brain. The latter finding is certainly in line with prior research that has reported no significant influence of HAC on Aβ [1]. It has been suggested that beer drinkers may have lower levels of Aβ in their brains, but the assessment of the extent of Aβ in the referred study was not based on the regional distribution of Aβ as in the current study but on the extent seen in one cortical section [28]. In contrast, recent animal studies have suggested that nonalcoholic liver steatosis promotes Aβ accumulation in the brain [44]. Consistent with the findings reported by Peng and colleagues, a clinical study assessing nonalcoholic fatty liver disease and plasma and imaging biomarkers of AD and vascular brain lesions suggested a link between midlife nonalcoholic fatty liver disease and dementia [33]. Noteworthy, the results obtained by us, Peng and colleagues in 2024 and Lu and colleagues in 2024 are not as such comparable as the methods differ significantly. We did see Aβ more frequently in the brains of subjects with liver inflammation compared to subjects without this liver alteration; however, the number of subjects with liver inflammation in our study was limited. Moreover, none of the assessed liver alterations influenced the severity of Aβ in the brain. This observation is in line with experimental studies suggesting that LP primarily influences glial cells (microglia) rather than Aβ accumulation [17]. Thus, based on our results, we cannot confirm the hypothesis that LP, i.e., steatosis, inflammation, or fibrosis, indeed influences the development of ADNC, i.e., HPτ and Aβ.

In 2020, in a report of the Lancet Commission discussion dementia prevention, intervention and care have listed alcohol consumption as a risk for dementia [32]. Noteworthy, references listed in this publication are various clinical studies as reports including neuropathological observations are scarce. In 2009, Clive Harper and in 2014 Suzanne de la Monte and Gillian Krill described and summarized alcohol-related brain damage and in both these publications the main alteration given is cell loss in various brain locations [16, 19]. We did not assess cell loss that eventually can be seen whereas we centered on protein alterations that are not described previously by others.

We created a Venn diagram to visualize the incidence of altered proteins in the brain among subjects with varying degrees of LP severity. Noteworthy, the number of subjects with concomitant alterations decreased significantly from 20% in those with mild LP to 3% in subjects with severe LP. Age at death is a significant factor in the observation of concomitant pathologies and it did not differ significantly between subjects with mild, moderate or severe LP. Thus, the difference cannot be attributed solely to the age of the subjects [3]. Noteworthy, dementia was registered in 30% of subjects with mild LP when compared with 13% of subjects with severe LP. Our observation, confirms that further studies on the brains of individuals with a history of HAC are warranted. Does severe LP indeed alter significantly some of the altered proteins associated with cognitive impairment?

Assessing alterations in peripheral tissues parallel to the evaluation of pathologies in the brain is of great interest. Previously, assessments of cardiovascular pathologies, as well as those of the kidney and pancreas, have been conducted to verify or deny existing associations between protein alterations in the brain and systemic diseases. Noteworthy, many of the proposed associations based on clinical or animal studies have not been confirmed, i.e., cardiovascular disease and diabetes do not appear to influence the extent of altered proteins in the brain, whereas these systemic diseases can certainly lead to vascular tissue damage in the brain [4, 21, 30].

The liver tissue was significantly affected in subjects with HAC; in individuals with severe LP, Alzheimer’s type II astrocytes were seen more frequently compared to controls. Astrocytes have been implicated in being of significance for Aβ processing [49, 53, 54]. In our 31 subjects with severe LP and frequently encountered Alzheimer type II astrocytes, however, no significant association was observed between the extent of LP and brain pathologies. The recently defined astrocytic alteration, ARTAG, i.e., glial HPτ pathology, was not influenced by LP. A significant association was, however, noted between Alzheimer’s type II astrocytes and ARTAG.

Studies integrating assessment of the brain and various peripheral organs are generally difficult to carry out. An autopsy is seldom performed [50]. Moreover, in many centers, especially when dealing with brain alterations, a brain-only autopsy is the preferred approach, i.e., the Netherlands Brain Bank. A long postmortem delay, often up to 240 h, is not unusual and can lead to tissue alterations that influence the assessment options. Thus, some cases may need to be excluded, particularly when peripheral organs such as the pancreas, gut and liver are included. Finally, and maybe most importantly, the visualizing techniques vary substantially. Some laboratories implement IHC, while others use in situ hybridization. Moreover, various antibodies and techniques (manual or various automated) are used. Thus, comparing results obtained by different laboratories is indeed difficult, if not impossible. Here, we assessed the brain pathology and the LP in a standardized manner (Tables 2 and 3) and noted that, as expected, the extent of all altered proteins in the brain correlated significantly and strongly with each other. Similarly, a significant correlation was noted between the different liver alterations. What has been demonstrated here emphasizes that human studies are both possible and informative, but they also present many pitfalls and challenges.

The overall sample size, whether it is 1000, 100, or 10 subjects, is certainly of significance. It has previously been reported that selection bias might alter the outcome [42]. Thus, here, we prefer to consider the results obtained from all 100 subjects to be reliable, while the outcomes when the cohort is separated based on gender or dementia are considered less reliable. When analyzing the data obtained from all 100 subjects with various extents of liver damage (LP ranging from 0 to 10), we noticed a negative correlation regarding the categorical data between liver steatosis and αS, we detected a significant negative correlation between the extent of liver steatosis and fibrosis and the extent of αS in the brain, we noticed a significant negative association between the observation of Alzheimer type II astrocytes and αS pathology in the brain, we detected a negative correlation regarding the categorical data between liver steatosis and Aβ and we noted no significant associations between LP and HPτ or TDP43.