The current study was the first large cohort study to explore the association between drinking behavior change patterns and HUA in the Chinese population. The findings showed that quitting drinking and continued drinking was associated with increased risks for HUA in the males, and the trends were more pronounced among those with mild to abstainer, mild to heavy, heavy to mild, and heavy to heavy drinking patterns. The magnitude of these independent associations increased further after adjusting for potential confounders. However, there was no association between the drinking patterns and HUA in the females. The rate of HUA was in sync with the estimations in our previously published meta-analysis that evaluated a whole population of 2,277,712 in China (15.5% vs 16.4%, respectively) [4].
Although the risk of HUA was lower in those who quit drinking compared to those with continued drinking, it still elevates the risk of HUA, compared to non-drinking. This result implies that early drinking could lead to an increased risk of HUA in males. The mechanism of decreased urate excretion has been implicated in the pathogenesis of alcohol-induced HUA. The study showed that HUA develops following conversion of alcohol to lactic acid, thus reducing uric acid excretion by competitively inhibiting uric acid secretion by the proximal tubule [23]. Faller et al. report that ethanol increases urate synthesis by enhancing the turnover of adenine nucleotides [24]. In addition, ethanol administration has been shown to increase the production of uric acid by enhancing the degradation of adenosine triphosphate to adenosine monophosphate, a uric acid precursor [25]. Our findings demonstrated that current heavy drinkers (drinking in 2009) had an increased risk of HUA in male participants (Supplementary figure 2), early mild or heavy drinkers (drinking in 1997) had increased risks of HUA (Figure 2). The consistent and significant association between mild to abstainer, mild to heavy drinking patterns, and HUA further validated the long-term effect of mild drinking patterns. We speculate that even mild alcohol intake could continuously decrease the glomerular filtration rate, which could promote the excretion of uric acid. Takashi et al. followed 8097 male workers for 8 years and showed that alcohol consumption at 2.5 gou/day (=ethanol 55 g/day) led to a distinct increase in the risk of HUA [26]. Baglietto et al. demonstrated that mortality curves were J-shaped (nadir at 9~12 g/day of alcohol consumption; the upper protective dose of 42~76 g/day) [27]. These findings showed that an average of 26 g/day (=18 SD*10/7 days, Figure 3) in 1997 or 16 g/day in 2009 (=11 SD*10/7 days, seen in Supplementary Figure 3) could cause a stable increase in the risk of HUA. The difference in threshold alcohol intake might be contributed to population heterogeneity (such as age, occupation, or health-related behaviors). As for the long-term effect of alcohol, our findings agreed with the Dietary Guidelines for Chinese Residents' report which showed that adult males should drink less than 25 g of alcohol per day [28].
Consistent with a single-center study in Liaoning of China [29], our findings demonstrated that alcohol consumption increased the risk of HUA only in males rather than females. It could be explained by the fact that the sample size of female drinkers was relatively small, thus leading to a low statistical power outcome. Besides, due to differences in androgen production, the ratio of uric acid to creatinine clearance is higher in women than in men [30, 31]. There is, therefore, a need for further studies to explore the mechanism underlying our findings. Of note, distinct risks of HUA in the three types of drinking were observed in our study. Liquor drinking at baseline led to a 1.8-fold increase in the risk of HUA compared with non-liquor drinking with a 1.1-fold risk per 200 ml (Figure 2c and d). A similar trend was observed in liquor drinkers in 2009 (Supplementary figure 2c and d). A 7-year cohort study (1988–1994) with 14,809 participants reported that increased SUA levels with increasing beer or liquor intake but not with increasing wine intake. However, the effect of ingested purine in beer on uric acid in blood might be sufficient to augment the HUA effect of alcohol in exerting a greater risk of gout than liquor or wine [32]. Previous studies showed that beer is the only alcoholic beverage with large purine content, which is predominantly guanosine. Guanosine is more readily absorbed than other nucleosides, nucleotides, or bases [33, 34]. Our data showed that drinking beer was marginally associated with HUA but without a dose-response relationship. Since beer contains large amounts of purines, it is feasible to speculate that the disparity in beer drinking in the male cohort could be due to the relatively small amount of beer consumption (an average of 2057 ml per week, data not shown). Because uric acid is considered an indicator for increased oxidative stress, polyphenols in wine with antioxidant properties might potentially play a role in mitigating the impact of alcohol on serum uric acid levels [35–37]. Furthermore, assessing the effect of drinking frequency in HUA showed that there was an increase in the magnitude of associations with increasing frequency of drinking [38, 39]. Thus, our findings provide a novel perspective that although the risk of HUA as a result of early drinking is lower than that associated with continued drinking, it still elevates the risk of HUA, as compared with the non-drinking.
Potential limitations of our study deserve comment. Firstly, our data lacked more than half of the variables on physical activity. To bridge this gap, we tried to adjust partly for total protein intake. In addition, since information on drinking behaviors was self-reported, inaccurate recall or under-reporting might have affected the results. Besides, our data failed to eliminate possible effects of underlying diseases and medications used for diseases such as uric-acid-lowering medication which might have affected the outcome.
Taken together, our study demonstrated that drinking behavior change patterns such as quitting and continued drinking are strongly associated with increased risks of HUA in males. The risks emanated from early drinking behaviors such as liquor drinking, high drinking frequency, and alcohol consumption. Although the risk of HUA in quitting drinking was lower than that in continued drinking patterns, it was positively associated with HUA. The long-term effect of early drinking behaviors on HUA could not be ignored.