Alcohol consumption in relation to cardiovascular and non-cardiovascular mortality in an elderly male Chinese population

There is consensus that alcohol intake at any quantity and of any class is detrimental [1]. Indeed, according to the Global Burden of Disease (GDB) study, alcohol intake was associated with 5.3, 3.3 and 4.1% of increased all-cause and cardiovascular mortality and cancers worldwide, respectively. The corresponding increases in all-cause mortality and cancers in the Chinese population were 4.1 and 6.2%, respectively [2, 3]. However, for now and probably for a long while in the future, alcohol drinking will continue as a major lifestyle factor in many populations. In China, alcohol intake is a habit mainly in men and rarely in women, and alcoholic beverages include beer and wine as in other populations and rice aperitif and white spirit typically in the Chinese population [4]. In addition, although regular drinking is prevalent, binge drinking is the predominant pattern in China, especially in liquor drinkers. Indeed, according to a recent national survey, 29.8% of male drinkers drank heavily once a week or more and 83.6% of them drank heavily occasionally [5].

We believe that liquor drinking, especially the binge drinking pattern, is particularly deleterious to health. One of the difficulties in drinking studies is that drinkers may stop drinking when they become unhealthy or have a severe disease diagnosed. Under such circumstances, the risk of drinking would be underestimated. We hypothesize that in the elderly past drinkers and liquor drinkers particularly have increased risks of cardiovascular and non-cardiovascular diseases. Such research requires a study population with a high proportion of drinkers including both past and current drinkers. We recently conducted an elderly population-based study in Shanghai, China. The proportion of alcohol intake was high in men [6‐9]. In the present study, we investigated the association between the quantity and type of alcoholic beverage consumption and the risk of cardiovascular and non-cardiovascular mortality in this elderly male Chinese population.

The key finding of our study was that past drinkers had a higher risk of all-cause and non-cardiovascular mortality. With regard to the alcoholic beverages, liquor/mix drinkers tended to have a higher risk of mortality, especially when compared with beer/wine drinkers.

Our observation that past drinkers had a higher risk of all-cause mortality is in keeping with the results of several previous studies [13, 14]. In an American prospective study, past drinkers (n = 24,904) had a higher risk of all-cause (HR 1.35, 95% CI 1.02–1.11) and cardiovascular (HR 1.27, 95% CI 1.17–1.37) mortality than lifetime abstainers (n = 76,869) [13]. In a Spanish study, past drinkers (n = 432) also had a higher risk of all-cause mortality than never drinkers (n = 941) both before (HR 1.49, 95%CI 1.03–2.14) and after exclusion of participants with functional limitations (n = 2665, HR 1.68, 95% CI 1.08–2.60) [14]. In the Prospective Urban Rural Epidemiology (PURE) study, however, past drinkers (n = 4255), compared with never drinkers (n = 74,685), had a significantly higher risk of all-cause mortality (HR 1.56, 95%CI 1.27–1.92) but non-significantly higher risk of fatal and nonfatal cardiovascular disease (HR 1.19, 95%CI 0.94–1.50) [15]. In addition, the risk in 1207 past drinker versus 1076 never drinkers was not significantly higher either for all-cause (HR 1.13, 95%CI 0.91–1.41) or cardiovascular mortality (HR 1.08, 95%CI 0.77–1.53) in a group of patients with a previous myocardial infarction, angina, or stroke enrolled in the UK Biobank Study [16]. The risk of past drinking might have been confounded by the existence of severe diseases.

One of the possible mechanisms for the higher disease and mortality risk of past drinkers is that past drinkers often stopped alcohol intake because of illnesses or health concerns [17]. Indeed, in our present study, past drinkers, compared with non-drinkers and current drinkers, had a significantly higher prevalence of hypertension and use of antihypertensive drugs and a slightly higher prevalence of diabetes mellitus.

In line with the results of several recent studies [14, 18, 19], our study did not show any beneficial effect of alcohol intake on cardiovascular disease mortality. In the abovementioned Spanish prospective study, neither light (1.43–20 g/d for men and < 10 g/d for women, n = 785) nor moderate (20–40 g/d for men and 10–24 g/d for women, n = 505) drinkers had a lower risk of mortality than never drinkers [14]. However, these recent observations are in contradiction to the cardiovascular protective effects observed in several earlier previous studies [13, 20‐22]. The latter observation might have been confounded by classifying past drinkers as abstainers. In a recent meta-analysis of 87 cohort studies (n = 3,998,626), after controlling for abstainer bias, light drinking (1.3–24.9 g/day) did not show any protective effect on the risk of mortality (relative risk 0.97, 95% CI 0.88–1.07) [18]. Similarly, in a cohort study in 24,029 US adults aged more than 50 years, low level of moderate drinking (< 7 drinks/week) had a similar risk of all-cause mortality as occasional drinkers (HR 1.02, 95% CI 0.94–1.11), whereas high level of moderate drinking (7 to < 14 drinks/week) had a significantly higher risk of all-cause mortality than occasional drinkers (HR 1.14, 95% CI 1.02–1.28) [19].

Abstaining might also have confounded the results of our study on the risks of heavy drinking. Heavy drinkers more likely stop drinking, because of illnesses or health concerns. In fact, current heavy drinkers had a higher systolic and diastolic blood pressure than non-drinkers, past drinkers and current light-to-moderate drinkers. The harmful effect of high blood pressure may appear with the longer-term follow-up.

Our observation on the lower risk of cardiovascular disease mortality in the current wine/beer drinkers is in accordance with the results of numerous previous studies [23‐25]. Indeed, in a recent cohort study (n = 114,970), wine (HR 0.88, 95% CI 0.69–1.13) and beer drinkers (HR 0.86, 95% CI 0.67–1.09), but not spirit or liquor drinkers (HR 0.97, 95% CI 0.81–1.16), tended to have lower risks of all-cause mortality [15]. Our study only included a small number of wine drinkers (n = 56). The number of events in wine drinkers was small (n = 1) and does not allow analysis in beer and wine drinkers separately.

Our study should be interpreted within the context of its limitations. First, we only collected information on the amount and type of alcoholic beverages at baseline but not during follow-up. Second, the quantity of alcohol consumption was self-reported. Over- or under-reporting is possible. Third, we did not collect information on the type of alcohol beverages or the lifetime drinking in past drinkers. We therefore were unable to perform analysis according to various alcoholic beverages and cumulative exposure to the risk in the past drinkers. Fourth, for the comparison between various alcoholic beverages, the number of events was small, leading to wide confidence intervals of the hazard ratios. Finally, because of the retrospective nature of the present analysis, sample size estimation was not performed prior to the analysis. Nonetheless, according to the observed proportion of alcohol intake, the present study may provide 84.9% of power to detect the observed association between alcohol intake and total mortality.

In summary, our study showed risks of mortality associated with past drinking and liquor drinking in men aged over 60 years of age in Shanghai, China. In fact, these two drinking factors are closely related. It is possible that high ethanol concentration of liquor drinking confers risks and when there are health concerns motivates abstaining from alcohol drinking. The latter hypothesis should be investigated in future studies of larger sample size and with information on the type of alcoholic beverages in past drinkers.