Cerebral Hemorrhage and Alcohol Exposure: A Review

Jialing Peng; Hongxuan Wang; Xiaoming Rong; Lei He; L Xiangpen; Qingy Shen; Ying Peng


Alcohol Alcohol. 2020;55(1):20-27. 

In This Article

Alcohol Consumption and Incidence of ICH

Most recent studies have confirmed the relationship between alcohol consumption and incidence of ICH as listed in Table 1 and Table 2. For the sake of comparability and analysis, the individual amount of intake is preferably standardized by a drink, and the exposure might be divided into several categories: light, moderate, high and heavy alcohol consumption (WHO, 2000; Larsson SC et al., 2016). The thresholds of the drinking categories mentioned above vary slightly in studies, which are attributed to the disparities of the ethanol content in a standard drink across countries. Divergent dose relationships between light-moderate alcohol consumption and ICH are delineated in different studies, while a definitely positive relationship between heavy drinking and ICH has been well accepted (Klatsky et al., 2002; Zhang et al., 2014; Larsson et al., 2016; Costa et al., 2018). An early population-based case-control study contrasted 467 incident stroke cases to 477 controls, with alcohol exposure assessed by a pre-coded questionnaire. The study demonstrated a strong relationship between ICH and heavy drinking, over 140 g/day, while no effect was revealed between ICH and moderate drinking, 1–30 g/day, which corresponded to the suggestion of a threshold phenomenon (Caicoya et al., 1999). After that, an international case-control study in 84 centers from 22 countries all over the world suggested a linear relationship between alcohol and ICH, after adjusting for age sex, and region. Alcohol intake was divided into two categories: 1–30 drinks per month and >30 drinks per month or binge drinker, both of which were associated with higher risk of ICH (O'Donnell et al., 2010). However, the early epidemiological studies neglected generally confounding factors including socioeconomic factors, stroke subtypes, geographical region, ethnic groups, misclassifications and underestimation of alcohol exposure and so on (WHO, 2000; O'Donnell et al., 2010) precluding conclusions on relationship between alcohol and ICH. For further understanding of the ascertained association, a recent study performed a meta-analysis which collected 27 prospective studies. Consumption was converted into standard drinks and four categories set as follows: light drinkers (<1 drink/day), moderate drinkers (1–2 drinks/day), high drinkers (>2–4 drinks/day) and heavy drinkers (>4 drinks/day). Another feature of this meta-analysis is that the reference groups were stratified into three categories: never drinkers, current non-drinkers and occasional drinkers. This meta-analysis indicated neither light nor moderate alcohol consumption was associated with higher incidence of ICH, while higher alcohol consumption had a non-statistically significant relation to an elevated risk of ICH, and heavy alcohol consumption was associated significantly with increased risk of ICH (Larsson et al., 2016), which was consistent with most previous studies (Klatsky et al., 2002; Zhang et al., 2014; Costa et al., 2018). Another study evaluated the relationship between alcohol consumption and the distribution subtypes of ICH, including putaminal hemorrhage (PH), thalamic hemorrhage (TH) and subcortical hemorrhage (SH), by multiple logistic regression analysis. The research indicated alcohol consumption was related to PH significantly but not a risk factor for TH or SH (Suzuki and Izumi, 2013). Recently, a multicenter case-control study involving patients of different races from 19 US sites analyzed the relationship between the location of ICH and chronic alcohol consumption. The subgroup demonstrated rare (<1 drink/month) and moderate (≥1 drink/month and ≤2 drinks/day) alcohol consumption decreased the risk of lobar and non-lobar ICH, whereas heavy (≥2 drinks/day) alcohol consumption was only related significantly to increased non-lobar ICH risk. In addition, the study validated the impact of race/ethnicity on alcohol-attributed ICH risk, and it was commendable to clarify the influence of alcohol on different ICH subtypes (Chen et al., 2017). In this regard, heavy alcohol intake predisposes deep ICH rather than lobar ICH, which is in line with other recent studies (Costa et al., 2018). Given that lobar ICH tends to be attributable to cerebral amyloid angiopathy (CAA), the findings indicated heavy alcohol exposure might have no significant influence on CAA-related ICH (Lavados et al., 2010; Ding et al., 2015). To date, studies to ascertain the association between alcohol consumption and CAA-related ICH are scarce and still needed. On account of various factors including age, sex, ethnic background, income and so on in different studies (van Asch et al., 2010), the relationship between low-middle alcohol consumption and ICH remains controversial. It is noteworthy that some abstainers quit drinking because of poor health, indicating risk of bias when abstainers are the reference group (Stockwell et al., 2016). One recent study suggested zero as the safest alcohol exposure level (95% uncertainty interval (UI), 0.0–0.8) based on pre-existing work (GBD 2016 Alcohol Collaborators, 2018) i.e. only 0 g of ethanol improves health (Burton and Sheron, 2018). All in all, it is well accepted that heavy (binge) drinking significantly increase the risk of ICH (Casolla et al., 2012; Costa et al., 2018). However, the underlying vasculopathy remains to be elucidated.