編者按：2018年8月28日，中國媒體《今日頭條》發佈了一篇文章：《一滴酒也別喝！世界權威醫學期刊已證實：沒有任何健康好處》，據稱，該文結論與其中所採用的數據，均來自於世界著名權威醫學期刊《柳葉刀》上的文章：《Alcohol use andburden for 195 countries and territories,1990–2016: a systematic analysis for the Global Burden of DiseaseStudy 2016》，筆者以前沒有看過《柳葉刀》這本期刊，也不知它是不是真發表了這麼一篇文章，不過在網上找到了該文章的英文原文，想必應該是真的，就想看看它的結論到底是不是真如《今日頭條》所言那樣，如果真是那樣，那麼文章用了何種研究方法，採用的數據樣本來自何處？本著這個目的，我們請英文翻譯節譯了該文，雖是節譯，但文章最關鍵部分都已經有了。不過由於時間倉促，在字句斟酌上難免不到位，甚至會有一些錯誤，敬請方家指正。
Alcohol use and burden for 195 countries and territories,
1990–2016: a systematic analysis for the Global Burden of Disease Study 2016
全球195個國家飲酒量及酒類消費的調查：1990---2016年飲酒與疾病的關係全面分析（節譯）
1
Summary
Background Alcohol use is a leading risk factor for death anddisability, but its overall association with health remains
complex given the possible protective effects of moderatealcohol consumption on some conditions. With our
comprehensive approach to health accounting within the GlobalBurden of Diseases, Injuries, and Risk Factors Study
2016, we generated improved estimates of alcohol use andalcohol-attributable deaths and disability-adjusted life-
years (DALYs) for 195 locations from 1990 to 2016, for bothsexes and for 5-year age groups between the ages of
15 years and 95 years and older.
概要
背景：酒精使用是導致死亡和殘疾的主要因素，鑑於適度飲酒對某些情況可能產生的保護作用，酒精與健康仍是息息相關的。
通過我們對全球疾病，傷害和風險因素研究中的綜合法，2016年，我們對1990年至2016年195個地點酒精使用和酒精引起的死亡和殘疾調整行了統計 - 男女性別和5歲以下年齡組15歲和95歲以上。
Methods Using 694data sources of individual and population-level alcohol consumption, along with592 prospective
and retrospective studies on the risk of alcohol use, weproduced estimates of the prevalence of current drinking,
abstention, the distribution of alcohol consumption amongcurrent drinkers in standard drinks daily (defined as 10 g
of pure ethyl alcohol), and alcohol-attributable deaths andDALYs. We made several methodological improvements
compared with previous estimates: first, we adjusted alcoholsales estimates to take into account tourist and
unrecorded consumption; second, we did a new meta-analysis ofrelative risks for 23 health outcomes associated with
alcohol use; and third, we developed a new method to quantifythe level of alcohol consumption that minimises the
overall risk to individual health.
方法
通過對個人和人群飲酒量的694個數據來源，以及592個預期關於酒精使用風險的回顧性研究，我們預估了當前飲酒的患病率。每日標準飲料中當前飲酒者的飲酒量分佈（定義為10克純乙醇）和酒精引起的死亡。我們做了幾個方法上的改進與之前的數據相比：首先，我們調整了酒精銷售預測，以考慮到遊客和未記錄的消費; 第二，我們對與23個相關健康結果的相對風險進行了新的分析，第三，我們開發了一種新的方法來量化酒精消費水平，最大限度地減少酒精攝入量對個人健康的整體風險。
Findings Globally,alcohol use was the seventh leading risk factor for both deaths and DALYs in2016, accounting for
2•2% (95% uncertainty interval [UI] 1•5–3•0) ofage-standardised female deaths and 6•8% (5•8–8•0) of age-
standardised male deaths. Among the population aged 15–49years, alcohol use was the leading risk factor globally in
2016, with 3•8% (95% UI 3•2–4•3) of female deaths and 12•2%(10•8–13•6) of male deaths attributable to alcohol
use. For the population aged 15–49 years, female attributableDALYs were 2•3% (95% UI 2•0–2•6) and male
attributable DALYs were 8•9% (7•8–9•9). The three leadingcauses of attributable deaths in this age group were
tuberculosis (1•4% [95% UI 1•0–1•7] of total deaths), roadinjuries (1•2% [0•7–1•9]), and self-harm (1•1% [0•6–1•5]).
For populations aged 50 years and older, cancers accountedfor a large proportion of total alcohol-attributable deaths
in 2016, constituting 27•1% (95% UI 21•2–33•3) of totalalcohol-attributable female deaths and 18•9% (15•3–22•6) of
male deaths. The level of alcohol consumption that minimisedharm across health outcomes was zero (95% UI 0•0–0•8)
standard drinks per week
.結果
在全球範圍內，酒精使用是2016年死亡和DALYs(疾病負擔的綜合性指標傷殘調整生命年)的第七大風險因素，佔齡標準化女性死亡的2.2％，男性死亡的6.8％。在15-49歲的人口中，酗酒是全球的主要風險因素。在15~49歲的人群中，3.8%的女性死亡要歸因於酒精，男性佔12.2％。結核病（1.4%）、交通事故（1.2%）和自我傷害（1.1%）是與酒精有關的主要死亡原因。
而對於年齡在50歲及以上的人來說，癌症是導致酒精相關死亡的主要原因，女性佔27. 1%，男性佔18.9%。
總的來說，與不飲酒的人相比，每天一杯會使相關健康風險增加0.5%，每天飲酒5杯風險則增加37％。
Interpretation Alcohol use is a leading risk factor forglobal disease burden and causes substantial health loss. We
found that the risk of all-cause mortality, and of cancersspecifically, rises with increasing levels of consumption, and
the level of consumption that minimises health loss is zero.These results suggest that alcohol control policies might
need to be revised worldwide, refocusing on efforts to loweroverall population-level consumption.
Funding Bill & Melinda Gates Foundation.
Copyright © 2018 The Author(s). Published by Elsevier Ltd.This is an Open Access article under the CC BY 4.0 license.
解釋
酒精使用是全球疾病負擔的主要風險因素，並導致嚴重的健康損失。我們
研究發現，全因死亡率和癌症的風險隨著消費水平的提高而上升
最小化健康損失的消費水平為零。這些結果表明，酒精控制政策可能會
需要在全球範圍內進行修訂，重新關注降低總體人口消費水平。
比爾和梅琳達蓋茨基金會資助。
版權所有©2018作者。由Elsevier Ltd.出版。這是CC BY 4.0許可下的Open Access文章
Introduction
Alcohol use has a complex association with health.
Researchers have recognised alcohol use as a leading risk
factor for disease burden, and studies link its consumption
to 60 acute and chronic diseases. 1–3 Additionally, some
research suggests that low levels of alcohol consumption
can have a protective effect on ischaemic heart disease,
diabetes, and several other outcomes. 4–6 This finding
remains an open question, and recent studies have
challenged this view by use of mendelian randomisation
and meta-analy
介紹
酒精使用與健康有著複雜的聯繫。
研究人員已將酒精使用視為主要風險
疾病負擔因素，研究發現其與60種急慢性疾病。 1-3節，一些
研究表明，飲酒量低可以對缺血性心臟病有保護作用，
4-6節說明這一發現仍然是一個懸而未決的問題。最近的研
通過使用孟德爾隨機化和元分析來證明這種觀點。
Determination of harm due to alcohol use is com-
plicated further by the multiple mechanisms through
which alcohol use affects health: through cumulative
consumption leading to adverse effects on organs and
tissues; by acute intoxication leading to injuries or
poisoning; and by dependent drinking leading to
impairments and potentially self-harm or violence. These
effects are also influenced by an individual’s consumption
volume and pattern of drinking. 2 Measuring the health
effects of alcohol use requires careful consideration of all
these factor
確定酒精使用造成的危害是通過多種機制進一步複雜化的。
哪些酒精使用影響健康：通過累積消費導致器官和器官的不良影響器官組織; 急性中毒導致傷害或中毒; 依賴飲酒導致損傷和潛在的自我傷害或暴力。這些效果也受個人消費的影響飲酒的量和模式影響。衡量酒精使用健康狀況的影響需要仔細考慮所有這些因素。
《柳葉刀》創刊號
《柳葉刀》於1823年由英國外科醫生，同時也是一名國會議員的湯姆·魏克萊（Thomas Wakley）所創刊，他以外科手術刀“柳葉刀”（Lancet）的名稱來為這份刊物命名，主要涉及糖尿病、腫瘤、傳染病等醫學領域的研究。
2
Research in context
Evidence before this study
Although researchers recognise alcohol use as a leading riskfactor
for premature death and disability, some evidence suggeststhat
low intake might have a protective effect on specificconditions
such as ischaemic heart disease and diabetes. Monitoring of
consumption behaviour is required to analyse the healtheffects of
alcohol use. Historically, researchers have relied onself-reported
survey data to estimate consumption levels and trends.However,
these data have systematic biases that make cross-country
comparisons unreliable. The Global Status Report on Alcoholand
Health, as well as previous iterations of the Global Burdenof
Diseases, Injuries, and Risk Factors Study, have sought toproduce
harmonised, cross-country comparisons of alcohol consumption
and its harms, by leveraging data on alcohol sales, theprevalence
of current drinking and abstention, and self-reports of
consumption amounts.
Added value of this study
In this analysis we improved available estimates of alcoholuse and
its associated health burden in five ways. First, weconsolidated
694 individual and population-level data sources to estimate
alcohol consumption levels among current drinkers. Second,
we developed a method to adjust population-level consumption
for alcohol consumed by tourists. Third, we improvedpre-existing
methods that account for unrecorded population-level
consumption. Fourth, we did a new systematic review and
meta-analysis of alcohol use and 23 associated healthoutcomes,
which we used to estimate new dose–response curves ofrelative
risk. Fifth, using the new relative risk curves and a newanalytical
method, we estimated the exposure of alcohol consumption that
minimises an individual’s total attributable risk.
Implications of all the available evidence
The total attributable burden of alcohol use was larger than
previous evidence has indicated and increases monotonically
with consumption. Based on weighted relative risk curves for
each health outcome associated with alcohol use, the level of
consumption that minimises health loss due to alcohol use is
zero. These findings strongly suggest that alcohol control
policies should aim to reduce total population-level
consumption. To potentially reduce the effects of alcohol use
on future health loss, there is a need for countries torevisit
their alcohol control policies and assess how they can be
modified to further lower population-level consumption.
研究背景
雖然研究人員認為對於過早死亡和殘疾酒精使用是一個主要的風險因素。一些證據表明
低攝入量可能對特定條件有保護作用，如缺血性心臟病和糖尿病。監測飲酒行為來分析健康的影響。從歷史上看，研究人員一直依賴於自我報道調查數據以估算消費水平和趨勢。然而，這些數據具有跨國的系統性偏見比較不可靠。關於酒精和飲酒的健康全球狀況報告，以及之前的全球負擔迭代疾病，傷害和風險因素研究，利用酒精銷售數據跨國比較它的危害，以及自我報告
消費金額，增加了本研究的價值。在此分析中，我們改進了可用的酒精使用估計值，廉價評估它對五種相關的健康的作用。首先，我們對694個人和人口年齡通過生活水平的數據來源進行估算其當前飲酒者的酒精消費水平。第二，我們開發了一種調整人口消費水平的方法引入用於遊客消費的酒精。第三，我們改進了預先存在的解釋未記錄的人口消費水平的方法。第四，我們做了一個新的系統評價和酒精使用的薈萃分析和23種相關的健康有關的結果，我們用它來估計相對的新劑量生成反應曲線。第五，使用新的相對風險曲線和新的分析方法，我們估計酒精消費的程度，最小化個人用量。所有證據說明酒精的壞處大於保護。以前的證據只能說明單調增加與消費。基於加權相對風險曲線與酒精使用相關的每種健康結果，最大限度地減少酒精使用造成的健康損失就是消費量為零。這些發現強烈暗示酒精控制政策應旨在降低總人口消費水平。減少酒精使用的影響，關於未來的健康，各國需要重新審視他們的酒精控制政策並評估他們的可能性，進一步降低人口消費水平。
Several studies have attempted to address these factors
to provide global estimates of alcohol consumption and its
associated health effects. The most comprehensive among
these studies have been WHO’s Global Status Report on
Alcohol and Health, as well as previous iterations of the
Global Burden of Diseases, Injuries, and Risk Factors
Study (GBD). 11–13 The present study aims to build upon pre-
existing work and to address several limitations found in
earlier research.
First, the available studies have assessed the risk of
alcohol use by relying on external meta-analyses, which do
not control for confounding in the selection of the
reference category within constituent studies. This
approach is problematic because of the so-called sick
quitter hypothesis, which emphasises the importance of
reference category selection in correctly assessing risk
among drinkers, along with other confounding study
characteristics such as survival bias. 8,14–17 Untilrecently,
most meta-analyses of alcohol consumption have not
controlled for the composition of the reference category.
Subsequently, assessments of harm relying on these
studies have been biased. We sought to resolve this issue
within our meta-analyses by including controls for
different reference categories and the average age of
participants.
Second, previous studies have used sales data to estimate
population-level alcohol stock. Researchers have noted the
benefit of using sales data instead of survey data for
quantifying alcohol stock available within a location. 18,19
However, sales data still have bias because of consumption
by tourists and unrecorded consumption from illicit sales,
一些研究試圖證明這些因素提供全球飲酒量及其消費量對健康影響。其中最全面的
這些研究是世界衛生組織的酒精和健康全球狀況報告，以及之前的迭代全球疾病負擔，傷害和風險因素研究（GBD）。 11-13頁本研究旨在建立預先現有的工作並解決了發現的一些侷限性
早期研究。首先，現有研究評估了風險依賴外部薈萃分析來酗酒中的類別。假設，強調了重要性
正確評估風險的參考類別，選擇在飲酒者中類加別，以及其他混淆特徵。 8,14-17頁直到最近，
大多數酒精消費的薈萃分析都沒有控制參考類別的組成。隨後，依靠這些評估做研究有偏見。我們試圖解決這個問題，通過包含控件來進行元分析不同的參考類別和平均年齡參與者。其次，以前的研究使用銷售數據進行估算人口級酒精庫存。研究人員已經注意到了使用銷售數據而不是調查數據的好處，量化一個地點內可用的酒精庫存。但是，消費，銷售數據仍然存在偏差因為包含
遊客和非法銷售的未記錄消費，家庭釀造和當地飲料。
home brewing, and local beverages. Without correction for
these factors, estimates relying on sales data can be biased
and lead to inaccurate cross-national comparisons. In the
current study, we adjusted the estimates of population-
level alcohol stock to account for the effects of tourism and
unrecorded consumption.
Third, previous studies have assumed zero as the
counterfactual exposure level that minimises harm. Within
a comparative risk assessment approach, a counterfactual
level of consumption that minimises harm is required
to estimate population attributable fractions (PAFs). 1
However, this counterfactual level needs to be estimated,
rather than assumed, given the complexities involved in
estimating the risk of alcohol use across outcomes. Relying
on this assumption can fail to capture any potential non-
linear effects between alcohol use and health. Our study
proposes a new method for the use of available evidence to
establish a counterfactual level of exposure across varied
relative risks, which provides tangible evidence for low-risk
drinking recommendations.
In the present study, we aimed to address these limita-
tions and provide the best available estimates of alcohol
use and the associated health burden. We estimated the
Research in context
Evidence before this study
Although researchers recognise alcohol use as a leading riskfactor
for premature death and disability, some evidence suggeststhat
low intake might have a protective effect on specificconditions
such as ischaemic heart disease and diabetes. Monitoring of
consumption behaviour is required to analyse the healtheffects of
alcohol use. Historically, researchers have relied onself-reported
survey data to estimate consumption levels and trends.However,
these data have systematic biases that make cross-country
comparisons unreliable. The Global Status Report on Alcoholand
Health, as well as previous iterations of the Global Burdenof
Diseases, Injuries, and Risk Factors Study, have sought toproduce
harmonised, cross-country comparisons of alcohol consumption
and its harms, by leveraging data on alcohol sales, theprevalence
of current drinking and abstention, and self-reports of
consumption amounts.
Added value of this study
In this analysis we improved available estimates of alcoholuse and
its associated health burden in five ways. First, weconsolidated
694 individual and population-level data sources to estimate
alcohol consumption levels among current drinkers. Second,
we developed a method to adjust population-level consumption
for alcohol consumed by tourists. Third, we improvedpre-existing
methods that account for unrecorded population-level
consumption. Fourth, we did a new systematic review and
meta-analysis of alcohol use and 23 associated healthoutcomes,
which we used to estimate new dose–response curves ofrelative
risk. Fifth, using the new relative risk curves and a newanalytical
method, we estimated the exposure of alcohol consumption that
minimises an individual’s total attributable risk.
Implications of all the available evidence
The total attributable burden of alcohol use was larger than
previous evidence has indicated and increases monotonically
with consumption. Based on weighted relative risk curves for
each health outcome associated with alcohol use, the level of
consumption that minimises health loss due to alcohol use is
zero. These findings strongly suggest that alcohol control
policies should aim to reduce total population-level
consumption. To potentially reduce the effects of alcohol use
on future health loss, there is a need for countries torevisit
their alcohol control policies and assess how they can be
modified to further lower population-level consumption.
Figure 1: Age-standardised prevalence of current drinking forfemales (A)
and males (B) in 2016, in 195 locations
Current drinkers are defined as individuals who reportedhaving consumed
alcohol within the past 12 months. ATG=Antigua and Barbuda.VCT=Saint
Vincent and the Grenadines. Isl=Islands. FSM=Federated Statesof Micronesia.
LCA=Saint Lucia. TTO=Trinidad and Tobago. TLS=Timor-Leste.
這些因素，依賴於銷售數據的估計可能存在偏差並導致不準確的跨國比較。
目前的研究，我們調整了人口估計數，酒精儲備水平，以及旅遊業的影響，未記錄的消費。
第三，先前的研究假設為零反事實暴露水平，最大限度地減少傷害。比較風險評估方法，反事實
需要最小化傷害的消費水平，估計人口歸因分數（PAFs）。但是，需要估計這種反事實水平，
考慮到所涉及的複雜性，而不是假設估計結果中酒精使用的風險。依託在這個假設下可能無法捕獲任何潛在的非酒精使用與健康之間的線性影響。我們的研究提出了一種利用現有證據的新方法
建立一個跨越不同的反事實暴露水平相對風險，為喝酒的低風險提供切實證據。
在本研究中，我們旨在解決這些限制並提供最佳的酒精估算值使用和相關的健康負擔。我們估計了研究背景，本研究之前的證據。雖然研究人員認為酒精使用是過早死亡和殘疾一個主要的風險因素，一些證據表明低攝入量可能對特定條件有保護作用如缺血性心臟病和糖尿病。監測
需要消費行為來分析健康的影響酒精使用。從歷史上看，研究人員一直依賴於自我報道調查數據以估算消費水平和趨勢。然而，這些數據具有跨國的系統性偏見比較不可靠。關於酒精和飲酒的全球健康狀況報告，以及之前的全球負擔迭代疾病，傷害和風險因素研究，都試圖生產酒精消費的統一，跨國比較通過利用酒精銷售數據，流行率，它的危害當前的飲酒，以及自我報告
消費金額。增加了本研究的價值在此分析中，我們改進了可用的酒精使用估計值
它有五種相關的健康負擔。首先，我們鞏固了694個人和人口水平的數據來源進行估算。
當前飲酒者的酒精消費水平。第二，我們開發了一種調整人口消費水平的方法
用於遊客消費的酒精。第三，我們改進了預先存在的解釋未記錄的人口水平的方法消費。第四，我們做了一個新的系統評價和酒精使用的薈萃分析和23種相關的健康結果，我們用它來估計相對的新劑量 - 反應曲線風險。第五，使用新的相對風險曲線和新的分析方法，我們估計酒精消費的暴露程度最小化個人的總歸屬風險。所有可用證據的含說明酒精使用的壞處大於以前的證據並且消費增加。基於加權相對風險曲線與酒精使用相關的每種健康結果，最大限度地減少酒精使用造成的健康損失是使消費量零。這些發現強烈暗示酒精控制政策應旨在降低總人口消費。減少有可能減少酒精使用的影響對未來的健康損失，各國需要重新審視他們的酒精控制政策並評估他們的可能性，健康以進一步降低人口消費水平。
圖1：當前女性飲酒的年齡標準化患病率（A）
和男性（B）在2016年，在195個地點
目前的飲酒者被定義為報告已經消費的個體
過去12個月內飲酒。 ATG =安提瓜和巴布達。 VCT =聖
文森特和格林納丁斯。 ISL =群島。 FSM =密克羅尼西亞聯邦。
LCA =聖盧西亞。 TTO =特立尼達和多巴哥。 TLS =東帝汶。
事實上，此前有些研究表明中低量飲酒可以降低全因死亡率。然而，這些研究有樣本量太小、混雜因素控制不足和計算相對風險的參考類別非最佳選擇等限制。最近一些採用孟德爾隨機化、彙集隊列研究和多變量薈萃分析的高質量研究，越來越多的證據發現飲酒對全因死亡率並無保護作用。
在這次的GBD研究數據中，我們通過圖片，可以得出更直接的結論。
3
prevalence of current drinking (having one or more
drinks in the past year); abstention from alcohol (having
no alcohol in the past year); the distribution of alcohol
consumption among current drinkers in standard drinks
daily; and the disease burden attributable to alcohol use,
in terms of deaths and disability-adjusted life-years
(DALYs). We produced these estimates for 195 locations
from 1990 to 2016, for both sexes and for 5-year age
groups between the ages of 15 years and 95 years and
older. We also did a new meta-analysis to assess the dose–
response risk of alcohol consumption for 23 outcomes.
Lastly, we estimated the level of alcohol consumption that
minimises an individual’s total attributable risk of anyhealth loss
prevalence of current drinking (having one or more
drinks in the past year); abstention from alcohol (having
no alcohol in the past year); the distribution of alcohol
consumption among current drinkers in standard drinks
daily; and the disease burden attributable to alcohol use,
in terms of deaths and disability-adjusted life-years
(DALYs). We produced these estimates for 195 locations
from 1990 to 2016, for both sexes and for 5-year age
groups between the ages of 15 years and 95 years and
older. We also did a new meta-analysis to assess the dose–
response risk of alcohol consumption for 23 outcomes.
Lastly, we estimated the level of alcohol consumption that
minimises an individual’s total attributable risk ofanyprevalence of current drinking (having one or more
drinks in the past year); abstention from alcohol (having
no alcohol in the past year); the distribution of alcohol
consumption among current drinkers in standard drinks
daily; and the disease burden attributable to alcohol use,
in terms of deaths and disability-adjusted life-years
(DALYs). We produced these estimates for 195 locations
from 1990 to 2016, for both sexes and for 5-year age
groups between the ages of 15 years and 95 years and
older. We also did a new meta-analysis to assess the dose–
response risk of alcohol consumption for 23 outcomes.
Lastly, we estimated the level of alcohol consumption that
minimises an individual’s total attributable risk of anyhealth loss.
家庭釀造和當地飲料一定要做為參考數據，依賴於銷售數據的估計可能存在偏差並導致不準確的跨國比較。目前的研究，我們調整了人口估計數 - 酒精儲備水平，包含旅遊業的影響未記錄的消費。第三，先前的研究假設為零，事實暴露，最大限度地減少傷害。比較風險評估方法，反事實需要最小化傷害的消費水平估計人口歸因分數（PAFs）。但是，需要估計這種反事實水平，考慮到所涉及的複雜性，而不是假設估計結果中酒精使用的風險。依託在這個假設下可能無法捕獲任何潛在的非酒精使用與健康之間的線性影響。我們的研究提出了一種利用現有證據的新方法建立一個跨越不同的反事實暴露水平，為低風險提供切實證據。在本研究中，我們旨在解決這些限制並提供最佳的酒精估算值使用和相關的健康負擔。
研究背景
本研究前的證據雖然研究人員認為酒精使用是對於過早死亡和殘疾一個主要的風險因素，一些證據表明低攝入量可能對特定條件有保護作用如缺血性心臟病和糖尿病。監測需要消費行為來分析健康的影響酒精使用。從歷史上看，研究人員一直依賴於自我報道調查數據以估算消費水平和趨勢。然而，這些數據具有跨國的系統性偏見比較不可靠。關於酒精和飲酒的全球狀況報告健康，以及之前的全球負擔迭代疾病，傷害和風險因素研究，都試圖生產酒精消費的統一，跨國比較通過利用酒精銷售數據，流行率，它的危害當前的飲酒，以及自我報告消費金額。增加了本研究的價值在此分析中，我們改進了可用的酒精使用估計值它有五種相關的健康負擔。首先，我們鞏固了694個人和人口水平的數據來源進行估當前飲酒者的酒精消費水平。第二，我們開發了一種用於統計遊客消費酒精水平的方法。第三，我們改進了預先存在的解釋未記錄的人口水平的方法消費。第四，我們做了一個新的系統評價和酒精使用的薈萃分析和23種相關的健康結果，風險。第五，使用新的相對風險曲線和新的分析方法，我們估計酒精消費的暴露程度最小化個人的總歸屬風險。所有可用證據的含義酒精使用的總歸責負擔大於以前的證據表明並且單調增加與消費。基於加權相對風險曲線與酒精使用相關的每種健康結果，水平最大限度地減少酒精使用造成的健康損失的消費量零。這些發現強烈暗示酒精控制政策應旨在降低總人口水平消費。有可能減少酒精使用的影響關於未來的健康損失，各國需要重新審視他們的酒精控制政策並評估他們的可能性修改以進一步降低人口消費水平。
Methods
Study design
This study follows the comparative risk assessment
framework developed in previous iterations of GBD. 20 In
the following sections, we summarise our methods and
briefly present innovations. A full explanation is available
in appendix 1. This study fully adheres to the Guidelines
for Accurate and Transparent Health Estimates Reporting
(GATHER) statement. 21
We estimated alcohol use exposure as grams of pure
ethanol consumed daily by current drinkers (which we
present here in terms of standard drinks daily, defined as
10 g of pure ethyl alcohol). We estimated relative risks by
dose in grams of pure ethyl alcohol, for each included
risk–outcome pair. We ascertained which cause and
injury outcomes to include by reviewing prospective and
observational studies of alcohol use, and then assessing
the causal association using Bradford-Hill’s criteria for
causation. 22 We included 23 outcomes, and the full list of
risk–outcome pairs, as well as the corresponding data
sources, are provided in appendix 1 (pp 52–140).
方法
本研究遵循比較風險評估在以前的GBD迭代中開發的框架。我們總結了我們的方法和簡要介紹創新。有完整的解釋
附錄1.本研究完全符合指南用於準確和透明的健康估計報告（GATHER）聲明。 2。1我們估計酒精使用暴露量為純淨克數當前飲酒者每天消耗的乙醇（我們在這裡以每日標準飲料的形式出現，定義為10克純乙醇）。我們通過估算相對風險劑量以克為單位的純乙醇，每種都包括在內風險 -結果對。我們確定了哪個原因和通過審查前瞻性和預後包括損傷結果酒精使用的觀察研究，然後評估使用Bradford-Hill標準的因果關聯因果關係。 22我們收錄了23項成果，以及完整的清單。風險 - 結果對，以及相應的數據
來源，見附錄1（第52-140頁）。
4
Data sources
We found sources that included indicators of current
drinking prevalence and alcohol consumed in grams per
day using the Global Health Data Exchange (GHDx) and
PubMed. 23 For the meta-analysis, we searched PubMed,
the GHDx, and references of previously published meta-
analyses. For our exposure estimates, we extracted
121 029 data points from 694 sources across all exposure
indicators. For our relative risk estimates, we extracted
3992 relative risk estimates across 592 studies. These
relative risk estimates corresponded to a combined study
population of 28 million individuals and 649 000 registered
數據源
我們發現包含當前指標的來源飲用流行率和酒精消耗量以克為單位使用全球健康數據交換（GHDx）和調查。 23對於薈萃分析，我們搜索了PubMed，GHDx，以及先前發表的meta-的參考文獻分析。對於我們的曝光估計，我們提取，來自694個來源的121 029個數據點指標。對於我們的相對風險估計，我們提取了592項研究的3992項相對風險評估。這些相對風險評估與綜合研究相對應人口2800萬，註冊人數649000各種結果的案例
5
cases of respective outcomes. We list all the included data
sources in appendix 1 (pp 52–140).
To estimate standard drinks consumed daily by current
drinkers, we followed the general approach used by Rehm
and colleagues. 18 We briefly explain this method here,
along with two methodological innovations to account
for bias in the sales model: an adjustment to account for
tourist consumption and an updated adjustment for
unrecorded consumption. A full explanation of this
approach is available in appendix 1 (pp 18–49).
To estimate exposure, we combined estimates of
population-level alcohol stock and individual-level alcohol
consumption to produce standard drinks consumed daily
among current drinkers and current drinker prevalence,
within a specific location, year, age group, and sex. We
started by estimating population-level alcohol stock inlitres
per capita from sales data, individual-level estimates of the
prevalence of current drinkers and abstainers from survey
data, and individual-level estimates of the amount of
alcohol consumed in grams per day from survey data.
Then, for a given location and year, we rescaled age-specific
and sex-specific estimates of individual-level consumption
so that they aggregated to the estimates of population-level
consumption. When surveys reported amount consumed
in terms of beverage types, we converted these data into
grams of pure ethanol using density equations and
assumptions of the average alcohol content by drink type
(appendix 1, p 50). Finally, we rescaled estimates of current
drinking and abstention so that, within a given location,
year, age group, and sex, the two estimates summed to one.
After we derived our model of population-level alcohol
stock from sales data, we controlled for sources of bias
that could arise from tourism and unrecorded con-
sumption not recorded in formal sales. To account for
tourist consumption, we computed an additive measure
for alcohol consumed abroad by domestic citizens and
subtractive measures for alcohol consumed domestically
by tourists. We extracted data on the number of tourists
by country of origin and destination from the World
Tourism Organization and used these data to obtain
estimates of total tourists, percentage of tourists by
location, and average duration of stay using a spatio-
temporal Gaussian process regression. 24 We combined
these estimates with measures of alcohol in litres per
capita by location, to calculate net amounts of total
population-level alcohol stock consumed by tourists or
domestic citizens travelling abroad.
To account for alcohol stock not captured within formal
alcohol sales data (ie, unrecorded consumption from
illicit production, home brewing, local beverages, or
alcohol sold as a non-alcohol product), we collated
estimates across published studies of the percentage of
total alcohol stock due to unrecorded consumption. We
sampled 1000 times from a uniform distribution with a
range between zero and the average of these collated
studies by location (sampling from the uncertainty
interval from each study, then averaging the draws) to
我們列出了所有包含的數據附錄1中的來源（第52-140頁）。估計當前每日消耗的標準飲料飲酒者，我們遵循Rehm使用的一般方法。 18我們在這裡簡要解釋這個方法，以及兩種方法論創新對於銷售模式的偏見：對帳戶的調整旅遊消費和最新調整未記錄的消費。對此的完整解釋方法見附錄1（第18-49頁）。為了估計明確，我們綜合估計人口級酒精和個人酒精消費每天生產標準飲料當前的飲酒者和目前的飲酒者流行率，在特定的位置，年份，年齡組和性別。我們首先通過估算人口級酒精含量來計算人均銷售數據，個人水平估計調查中當前飲酒者和戒酒者的流行程度數據，以及個人數量的估計數量。從調查數據中每天以克為單位消耗的酒精。然後，對於給定的位置和年份，我們重新調整了特定年齡個人消費的性別特定估計，這樣他們就可以彙總到人口水平的估計值消費。當調查報告消費量時在飲料類型方面，我們將這些數據轉換為克純乙醇使用密度方程和飲料類型的平均酒精含量的假設（附錄1，第50頁）。最後，我們重新調整了當前的估計值飲酒和棄權，以便在特定地點，年齡組和性別，這兩個估計總和為一。在我們得出我們的人口級別酒精模型之後根據銷售數據庫存，我們控制了偏差來源可能來自旅遊和未記錄的沒有在正式銷售中記錄的消費。要佔旅遊消費，我們計算了一個附加措施國內公民和國外消費的酒精減去國內飲酒量的措施及遊客。我們提取了有關遊客數量的數據，來自世界的原籍國和目的地國旅遊組織並使用這些數據來獲取估計遊客總數，遊客百分比使用空間的位置和平均逗留時間時間。 24我們合併了這些估算採用酒精每升的量度按地點劃分人均，計算總淨額遊客或消費者消費的人口級酒精包括出國旅遊的國內公民。計算未在正式範圍內捕獲的酒精庫存，酒精銷售數據（即未記錄的消費量）非法生產，家庭釀造，當地飲料，或作為非酒精產品銷售的酒精），我們整理對公佈的百分比研究進行估計由於未記錄消費而導致的酒精總量。我們從均勻分佈中採樣1000次範圍在零和這些整理的平均值之間地點研究（從不確定性中抽樣）每次研究的間隔，然後平均繪製圖）對沒有被記錄可能的總存量進行保守估計。
6
generate a conservative estimate of the total stock likely
to be unrecorded. We used a conservative approach
because of the wide heterogeneity in both the methods
and estimates within included data sources. We provide
estimates of these percentages in appendix 1 (pp 46–49).
Systematic review and meta-analysis
We did a new systematic review for each associated
outcome to incorporate new findings on risk and to
improve upon existing approaches. This strategy allowed
us to systematically control for reference category con-
founding in constituent studies across associated out-
comes. We provide the search strategy, search diagrams,
dose–response curves for each included outcome, and
references for each outcome in appendix 1 (pp 57–146).
Drawing from our systematic review, we did a meta-
analysis of risk outcomes for alcohol use. For each
outcome, we estimated the dose–response relative risk
curve using mixed-effects logistic regression with non-
linear splines for doses between 0 and 12•5 standard
drinks daily. We selected 12•5 standard drinks daily as a
cutoff point given the absence of available data beyond
this range. We present additional details of the model in
appendix 1 (pp 51–138). We tested the significance of
including a study-level confounding variable on the
composition of the reference category (eg, whether former
drinkers were included in the abstainer category or not).
When found to be significant, this variable was included
as a predictor within the model, which was the case for
ischaemic heart disease, ischaemic stroke, and diabetes.
Using our dose–response curves, we estimated the
consumption level that minimises harm, which is
defined in the comparative risk assessment approach as
the theoretical minimum risk exposure level (TMREL).
We chose a theoretical minimum on the basis of a
weighted average relative risk curve across all attributable
outcomes. We constructed weights for each risk outcome
based on the respective global, age-standardised DALY
rate per 100 000 in 2016 for both sexes. Our TMREL was
the minimum of this weighted all-attributable outcome
dose–response curve.
Attributable burden due to alcohol use
We calculated PAFs using our estimates of exposure,
relative risks, and TMREL, following the same approach
taken within the GBD studies. 20 For alcohol-use disorders,
which are by definition fully attributable, we assumed a
PAF of 1. 24 Following this calculation, we multiplied PAFs
by outcome-specific estimates of deaths and DALYs and
summed these across outcomes to calculate the total
attributable burden in specific locations. We aggregated
both exposure and burden results at the global level and
have presented them by quintile of the Socio-demographic
Index (SDI). SDI is a summary measure of overall
development, based on educational attainment, fertility,
and income per capita within a location. Locations
categorised by SDI quintile are found in appendix 1
我們採用了保守的方法因為這兩種方法都存在很大的異質性並在包含的數據源中進行估算。我們提供
附錄1中對這些百分比的估計（第46-49頁）。系統評價和薈萃分析我們對每個相關的人進行了新的系統評價結果將新發現納入風險和改進現有方法。這個策略允許我們系統地控制參考類別在相關的外部成分研究中成立。我們提供搜索策略，搜索圖表，每個納入結果的劑量 - 反應曲線，和附錄1中每個結果的參考文獻（第57-146頁）。根據我們的系統評價，我們做了一個分析酒精使用的風險結果。對於每一個結果，我們估計了劑量反應的相對風險曲線使用混合效應邏輯，每天喝劑量在0到12•5標準之間的線性樣條。我們每天選擇12•5標準飲料截止這個範圍作為可用數據，。我們提供了該模型的其他細節附錄1（第51-138頁）。我們測試了它的重要性包括研究級別的混雜變量參考類別的組成（例如，是否為前者飲酒者是否包括在戒酒者類別中。當發現時，包括該變量作為模型中的預測，情況就是如此缺血性心臟病，缺血性中風和糖尿病。使用我們的劑量 - 反應曲線，我們估計了最小化傷害的消費水平，即在比較風險評估方法中定義為理論最小風險暴露水平（TMREL）。我們在a的基礎上選擇了理論最小值所有歸屬的加權平均相對風險曲線結果。我們為每個風險結果構建了權重基於各自的全球年齡標準化DALY。2016年兩性均為每10萬人的比率。我們的TMREL是此加權全部歸因結果的最小值劑量 - 反應曲線。酒精使用造成的歸因負擔我們使用我們的暴露估計來計算PAF，在GBD研究中相對風險和TMREL遵循相同的方法進行。 20對於酒精使用障礙，根據定義，完全可歸因於我們假設PAF為1. 24在此計算之後，我們將PAF乘以通過結果特定的死亡和DALYs估計將這些結果相加以計算總數特定地點的可歸責。我們彙總了在全球範圍內產生量。通過社會人口的五分之一提出了它們指數（SDI）。 SDI是總體的總結度量發展，基於教育程度，生育率，一個地方的人均收入。地點按SDI五分位數分類見附錄1（第8-12頁）。
7
(pp 8–12). 25 We also constructed age-standardised values
of all estimates, using the same age weights as those used
in the GBD standard population.
We made one adjustment to road injury PAFs to estimate
how much burden occurred to others because of alcohol
use by another individual. We based this adjustment on
data from the US Fatality Analysis Reporting System
(FARS), which includes the average number of deaths in
automobile accidents involving alcohol and the percentage
of those deaths distributed by age and sex. We multiplied
age-specific and sex-specific alcohol-attributable and road-
injury-attributable DALYs by the average number of
fatalities, given the driver’s age and sex. We then re-
distributed these attributable DALYs according to the
FARS-derived probabilities that a population by age and
sex would be involved in a road injury, given the exposed
driver’s age and sex. Because of data availability, we
assumed that locations outside the USA would follow a
similar pattern to what we estimated with FARS. After
redistributing the attributable DALYs, we derived PAFs
again by dividing the redistributed attributable DALYs by
total DALYs within specific demographics.
Uncertainty analysis
For all steps, we calculated uncertainty for estimation of
exposure, attributable deaths, and DALYs by taking 1000
draws from the data’s uncertainty due to sampling error
and modelling uncertainty arising from hyper-parameter
selection and parameter estimation. We then used these
draws throughout the entire modelling process. When
reporting uncertainty intervals, we present the 2•5th and
97•5th percentiles of the draws.
Role of the funding source
The funders of the study had no role in study design,
data collection, data analysis, data interpretation, or
writing of the report. The corresponding author had full
access to all the data in the study and had final
responsibility for the decision to submit for publication.
Results
Global, regional, and national trends in alcohol
consumption
In 2016, 32•5% (95% uncertainty interval [UI] 30•0–35•2)
of people globally were current drinkers. 25% (95% UI
23–27) of females were current drinkers, as were
39% (36–43) of males (appendix 2). These percentages
corresponded to 2•4 billion (95% UI 2•2–2•6) people
globally who were current drinkers, with 1•5 billion
(1•4–1•6) male current drinkers and 0•9 billion (0•8–1•0)
female current drinkers (appendix 2, pp 2–1994). Globally,
the mean amount of alcohol consumed was 0•73 (95% UI
0•68–0•78) standard drinks daily for females and
1•7 (1•5–1•9) standard drinks daily for males.
The prevalence of current drinking varied considerably
by location (figure 1). Prevalence was highest for high
SDI locations, where 72% (95% UI 69–75) of females and
我們還構建了年齡標準化的價值觀所有估計數，使用相同的年齡在GBD標準人口中。我們對車禍害PAF進行了一次調整估算，因酒精而給別人帶來了多少負擔。我們基於此調整來自美國致命分析報告系統的數據（FARS），其中包括平均死亡人數涉及酒精的汽車事故和百分比按年齡和性別分配的死亡人數我們成倍增加，特定年齡和性別特定的酒精歸因和道路 - 傷害歸因DALYs的平均數死亡人數，同時考慮到司機的年齡和性別。然後我們重新分配這些可歸因的DALYs，FARS導出的概率，即按年齡和年齡劃分的人口數量。由於數據可用性，我們假設美國境外的地點會跟隨與我們用FARS估計的模式類似。後重新分配歸因DALY，我們得出了PAF再次將重新分配的歸屬DALY除以特定人口統計數據中的總DALYs。
不確定性分析
對於所有步驟，我們計算了估計的不確定性，歸因於死亡和DALYs，取1000由於採樣誤差導致數據的不確定性和超參數引起的不確定性建模選擇和參數估計。然後我們使用了這些在整個建模過程中繪製。什麼時候報告不確定區間，我們提出2•5和97•平局的第5個百分位數。資金來源的作用該研究的資助者在研究設計中沒有任何作用，數據收集，數據分析，數據解釋或撰寫報告。作者訪問研究中的所有數據並進行最終決定決定提交出版。
結果
全球，區域和國家酒精趨勢消費
2016年，32•5％（95％不確定區間[UI] 30•0-35•2）全球人都是當前的飲酒者。 25％（95％UI
23-27）女性當前是飲酒者39％（36-43）的男性（附錄2）。這些百分比相當於2•40億（95％UI 2•2-2•6）的人全球目前的飲酒者，擁有15億美元（1•4-1•6）男性當前飲酒者和0.90億（0•8-1•0）女性當前飲酒者（附錄2，第2-1994頁）。全球範圍內，平均消耗的酒精量為0.73（95％UI0•68-0•78）女性和女性每日標準飲料男性每日1•7（1•5-1•9）標準飲品。目前飲酒的流行程度差異很大按位置（圖1）。患病率最高SDI位置，72％（95％UI 69-75）的女性和
（第8-12頁）。總結
從這份數據龐大的嚴謹報告中可以得出，飲酒增加全因死亡率，儘管酒精可以帶來些微的保護作用，但被其他風險中和抵消了。在50歲以後，癌症是酒精相關死亡率最重要的因素。
83％（80-85）的男性是當前的飲酒者（地點附錄2中提供了每個SDI五分位數，第8-12頁）。飲酒流行率在低至低水平時最低中間SDI位置，其中8•9％（95％UI 6•6-9•7）女性和20％（17-22）的男性是當前的飲酒者。在SDI五分位數中，女性消耗的酒精較少與男性相比，這種差距的大小隨之減小更高的SDI水平。例如，我們發現尼泊爾的女性和男性之間存在差異，目前只有1•5％（95％UI 1•2-1•9）的女性飲酒。2016年的飲酒者與男性21％（17-25）。相反，許多高SDI位置都相就小。例如，我們發現瑞典的差異很小，其中86％（UI為95％）（84-88）女性和87％（85-89）的男性當前的飲酒者消費的人口平均值差異好小。2016年當前飲酒者每天也有很大差異按地點和性別劃分（圖2）。高SDI位置有每日消費的標準平均值最高每天消耗1•9（95％UI 1•3-2•7）在女性中，男性中有2•9（2•0-4•1）。低SDI位置的男性平均值最低每日消耗1•4（0•6-2•4）標準飲料中低端SDI位置的平均值最低對於女性每天消耗標準為0•3（0•1-0•6）
。
酒精引起的死亡和死亡的全球模式
疾病
2016年，有2,800萬人死亡（95％UI 2•4-3•3）歸因於酒精的使用。這相當於2•2％（95％UI 1•5-3•0）總年齡標準化死亡人數女性和男性中的6•8％（5•8-8•0）。整體疾病，酒精使用導致1•6％（95％UI女性中2016年全球DALY總數的1•4-2•0）男性中有6•0％（5•4-6•7）。在全球範圍內，酒精使用被評為第七大風險因素，2016年過早死亡和殘疾GBD研究中的其他風險因素。人口年齡在15-49歲之間，酒精使用率居首位風險可歸因為疾病全球風險因素，引起8.9％（95％UI 7•8-9•9）的歸因DALYs男性和女性的2•3％（2•0-2•6）。人口年齡在15-49歲之間，3•8％（95％UI 3•2-4•3）的人口女性死亡和12•2％（10•8-13•6）的男性死亡可歸因於酒精使用。酒精使用的總負擔和與酒精使用相關的原因比例各不相同。按性別，年齡和SDI五分之一（圖3;附錄2，第1997-2186頁）。從絕對意義上講，酒精可歸因女性的年齡小於男性（圖3）。對於女性來說，酒精可歸因於負擔隨著年齡的增長而增加，而男性則增加了負擔直到55-65歲之間，之後歸屬減少。女性，尤其是高SDI地點，有一些保護作用，缺血性心臟病和60歲以上的糖尿病年齡段最明顯。對於男性，只有高SDI和低SDI位置對缺血性心臟病有明顯的保護作用，但效果與總量相比較小。對於男性和女性，健康結果在整個生命中改變可歸因的負擔 - 跨度（圖3）。歸因的三大主要原因該年齡組的死亡人數為肺結核（1•4％[95％UI]1•0-1•7]總死亡人數），道路傷害（1•2％[0•7-1•9]），和自殘（1•1％[0•6-1•5]）。對於年齡較大的女性15-49歲，酒精使用障礙構成最大SDI中可歸因負擔的比例五分之一;主要的例外是塊莖結核病佔應占比例最大。在這個年齡段，運輸傷害和酒精使用障礙是主要的中高級男性可歸因的原因SDI五分位數;適用於中低SDI和低SDI五分位數是結核病的主要原因。、超過50歲，總歸因的原因SDI五分之一的負擔變得更加複雜。對於50歲及以上的人群，癌症佔佔酒精總死亡人數的很大一部分
2016年，構成27•1％（95％UI 21•2-33•3）的總酒精引起的女性死亡人數和18•9％（15•3-22•6）酒精引起的男性死亡。在高SDI國家，癌症是其主要來源兩性的歸因負擔。在低SDI國家，結核病是造成負擔的主要原因。對於男女兩性，其次是肝硬化和其他慢性病肝病。
高可歸因負擔的概況
男性主要由缺血性中風組成和出血性中風，其次是肝病在所有SDI五分位數，出血性中風和高血壓性心臟病是最大的來源。80歲及以上女性的多發這些疾病。對於男人來說這個年齡段，構成是類似疾病。
8
minimum relative risk of 0•86 (0•80–0•96) for men and
0•82 (0•72–0•95) for women, occurring at 0•83 standard
drinks daily for men and 0•92 standard drinks daily for
women. We found no significant difference in relative
risk curves for ischaemic heart disease or diabetes
when estimating the curves by age. For all other out-
comes, including all cancers, we found that relative risk
monotonically increased with alcohol consumption
(appendix 2, pp 57–146).
In estimating the weighted relative risk curve, we
found that consuming zero (95% UI 0•0–0•8) standard
drinks daily minimised the overall risk of all health
loss (figure 5). The risk rose monotonically with
increasing amounts of daily drinking. This weighted
relative risk curve took into account the protective effects
of alcohol use associated with ischaemic heart disease
and diabetes in females. However, these protective
effects were offset by the risks associated with cancers,
which in creased monotonically with consumption. In a
sensitivity analysis, where we explored how the weighted
relative risk curve changed on the basis of the choice of
weights for various health outcomes, the curve changed
signifi cantly only in settings where diabetes and
ischaemic heart disease comprised more than 60% of
total deaths in a population.
Discussion
In 2016, alcohol use led to 2•8 million deaths and was the
leading risk factor for premature death and disability
among people aged 15–49 years, with nearly 9% of all
attributable DALYs for men and more than 2% for
women. Our findings indicate that alcohol use was
associated with far more health loss for males than for
females, with the attributable burden for men around
three times higher than that for women in 2016. By
evaluating all associated relative risks for alcohol use, we
found that consuming zero standard drinks daily
minimises the overall risk to health.
Previous research has analysed all-cause risk due to
alcohol use by either investigating all-cause risk in
particular cohorts and survey series, or through meta-
analyses of those studies. 26,27 Past findings subsequently
suggested a persistent protective effect for some low or
moderate levels of alcohol consumption on all-cause
mortality. However, these studies were limited by
small sample sizes, inadequate control for confounders,
and non-optimal choices of a reference category for
calculating relative risks. More recent research, which
has used methodologies such as mendelian randomis-
ation, pooling cohort studies, and multivariable adjusted
meta-analyses, increasingly shows either a non-signifi-
cant or no protective effect of drinking on all-cause
mortality or cardiovascular outcomes. 7,14,28 Our results on
the weighted attributable risk are consistent with this
body of work. Taken together, these findings emphasise
that alcohol use, regardless of amount, leads to health
loss across populations. Although we found some
protective effects for ischaemic heart disease and
diabetes among women, these effects were offset when
overall health risks were considered—especially because
of the strong association between alcohol consump-
tion and the risk of cancer, injuries, and communic-
able disease. These findings stress the importance of
assessing how alcohol use affects population health
across the lifespan.
Evaluating attributable burden across SDI quintiles
revealed the magnitude by which outcomes of alcohol
use differ and how total attributable burden relates to
increasing SDI. Our results indicate that alcohol use and
its harmful effects on health could become an increasing
challenge amid gains in SDI. Given that most low and
low-to-middle SDI settings currently have lower average
alcohol consumption than high-to-middle SDI settings,
it is crucial for decision makers and government agencies
to enact or maintain strong alcohol control policies today
to prevent the potential for rising alcohol use in the
future. Effective policies now could yield substantial
population health benefits for years to come.
男性和女性的最低相對風險為0•86（0•80-0•96）女性0•82（0•72-0•95），。男士每日飲酒發生率為0.83，女性每日飲用0•92。我們發現相對沒有顯著差異缺血性心臟病或糖尿病的風險曲線按年齡估算曲線時對於 - ，包括所有癌症，我們發現相對風險單調增加飲酒量（附錄2，第57-146頁）。在估計加權相對風險曲線時，我們發現消耗零（95％UI 0•0-0•8）標準每飲量可最大限度地降低所有健康的總體風損失（圖5）。風險單調上漲，日常飲酒量增加。這加權了相對風險曲線考慮了保護作用與缺血性心臟病相關的酒精使用和女性的糖尿病。但是，這些保護性癌症相關風險抵消了這些影響，
隨著消費單調增加。有一個敏感性分析，我們在哪裡探討加權相對風險曲線在選擇的基礎上發生了變化，加上了各種健康結果的權重，人口中的總死亡人數曲線發生變化，不只在糖尿病和糖尿病的環境中有影響，缺血性心臟病佔60％以上。
討論
在2016年，酒精使用導致了2,800萬人死亡，並且是過早死亡和殘疾的主要風險因素在15-49歲的人中，有近9％的人男性的歸因DALYs為2％以上婦女。我們的研究結果表明，酒精使用是與男性相比，女性健康損失遠遠大於男性，對男性來說可歸因於負擔比2016年的女性高出三倍。我們評估了酒精使用的所有相關相關風險。發現每天消費零標準酒精最大限度地降低整體健康風險。
以前的研究已經分析了由此產生的全因風險通過調查全因風險來使用酒精特別的同期群和調查系列，或通過meta-這些研究的分析。 26,27的調查結果顯示一些低或者持久的保護作用全因飲酒量適中死亡。然而，這些研究受到限制樣本量小，對混雜因素控制不足，參考類別的非最佳選擇，並計算了相對風險。最近的研究，其中使用了諸如孟德爾隨機的方法 - 組合，隊列研究和多變量調整薈萃分析，越來越多地顯示出非顯著性飲酒對全因無影響或對心血管無保護作用。7,14,28的結果加權歸屬風險與此一致。總之，這些發現強調了這一點酒精的使用，無論數量多少，都會導致健康受損，人口減少。雖然我們發現了一些缺血性心臟病的保護作用，女性糖尿病，使這些保護這些影響被抵消了。考慮了整體健康風險 - 特別是因為酒精與癌症之間的強關聯和。這些發現強調了重要性評估酒精使用如何影響人口健康，人類整個生命週期。評估SDI五分位數的可歸因負擔揭示了酒精結果的大小使用不同以及可歸因負擔總額如何相關增加SDI。我們的結果表明，酒精的使用和它對健康的有害影響可能會增加在SDI的收益中挑戰。鑑於最低和中低SDI設置目前的平均值較低酒精消耗量高於中高SDI設置，這對政府機構至關重要，今天制定或維持強有力的酒精控制政策。
9
Our results point to a need to revisit alcohol control
policies and health programmes, and to consider
recommendations for abstention. In terms of reducing
population-level alcohol use, WHO provides a set of best
buys—policies that provide an individual year of healthy
life at less than the cost of the average individual
income. 29 Governments should consider how these
recommendations can be implemented within their local
contexts and broader policy platforms, including excise
taxes on alcohol, controlling the physical availability of
alcohol and the hours of sale, and controlling alcohol
advertising. Any of these policy actions would contribute
to reductions in population-level consumption—an
important step toward decreasing the health loss
associated with alcohol use.
Failing to address harms from alcohol use, particularly
at high levels of consumption, can have dire effects on
population health. The mortality crisis in Russia is a
striking example, where alcohol use was the primary
culprit of increases in mortality starting in the 1980s and
led to 75% of deaths among men aged 15–55 years. 30
Current global trends—namely, population ageing—
portend a growing toll of the alcohol-attributable burden
in the absence of policies, particularly since many cancers
disproportionately affect older individuals. Consequently,
low-to-middle SDI countries could benefit from policy
action today to keep alcohol consump tion low and
prevent greater health loss in the future. High and
high-to-middle SDI locations need to consider stronger
alcohol reduction policies, such as those recommend-
ed by WHO, in an effort to reduce population-level
consumption.
Our results should be interpreted within the context of
the study’s limitations. First, our consumption estimates
might not fully capture illicit production or unrecorded
consumption given our use of sales data in estimation.
We have sought to adjust for consumption beyond sales
data, but given the heterogeneity of these estimates it is
likely that additional methodological refinements are
necessary to improve the quantification of unrecorded
consumption. Second, drinking patterns within a year
are assumed to be consistent; however, past work shows
that drinking patterns, rather than average levels of
consumption such as standard daily drinks, might be
related to different levels of risk and harm. Unfortunately,
the data requirements for assessment of such drinking
patterns by age, sex, and location far exceed what is
currently available. For instance, few prospective studies
quantify the effects of drinking patterns and average
levels of consumption in tandem, a requirement for
correctly assessing the risk of alcohol-attributable
outcomes. Third, the data used to estimate motor vehicle
harm caused to others from alcohol use were only
available for the USA (ie, FARS data). Although it is
unlikely that the patterns observed in FARS are drastically
different from those of other locations (appendix 1,
pp 141–144), this assumption needs to be tested with
more location-specific estimates. Fourth, we were unable
to find robust data about the harm caused to others
from alcohol-attributable interpersonal violence, a major
potential source of health loss. More retrospective studies
are needed to assess the harm to others caused through
an individual’s alcohol use. 30 Fifth, consumption for
populations younger than 15 years was not assessed
because of data sparseness on alcohol use for these age
groups. In the absence of such data, potential approaches
to address this limitation, such as assuming consumption
patterns of older age groups or trying to extrapolate past
levels of alcohol consumption, are likely to introduce
additional bias or error. More research on youth drinking
and the associated risk is required to estimate alcohol-
attributable burden for this age group. Last, we sought to
quantify the risk of alcohol use only for outcomes with
evidence meeting the criteria for the comparative risk
assessment approach of GBD studies. However, there are
additional outcomes, such as dementia and psoriasis,
for which accumulating evidence suggests that alcohol
use might be a risk factor. 31–33 In combination, these
limitations suggest that our results are likely to under-
estimate both the health risks and overall attributable
burden of alcohol use.
Conclusion
Alcohol use is a leading risk factor for disease burden
worldwide, accounting for nearly 10% of global deaths
among populations aged 15–49 years, and poses dire
ramifications for future population health in the absence
of policy action today. The widely held view of the health
benefits of alcohol needs revising, particularly as
improved methods and analyses continue to show how
much alcohol use contributes to global death and
disability. Our results show that the safest level of
drinking is none. This level is in conflict with most
health guidelines, which espouse health benefits
associated with consuming up to two drinks per day.
Alcohol use contributes to health loss from many causes
and exacts its toll across the lifespan, particularly among
men. Policies that focus on reducing population-level
consumption will be most effective in reducing the
health loss from alcohol use.
GBD 2016 Alcohol Collaborators
Max G Griswold, Nancy Fullman, Caitlin Hawley, NicholasArian,
Stephanie R M Zimsen, Hayley D Tymeson, Vidhya Venkateswaran,
Austin Douglas Tapp, Mohammad H Forouzanfar, Joseph S Salama,
Kalkidan Hassen Abate, Degu Abate, Solomon M Abay,
Cristiana Abbafati, Rizwan Suliankatchi Abdulkader, ZegeyeAbebe,
Victor Aboyans, Mohammed Mehdi Abrar, Pawan Acharya,
Olatunji O Adetokunboh, Tara Ballav Adhikari, Jose C Adsuar,
Mohsen Afarideh, Emilie Elisabet Agardh, Gina Agarwal,
Sargis Aghasi Aghayan, Sutapa Agrawal, Muktar Beshir Ahmed,
Mohammed Akibu, Tomi Akinyemiju, Nadia Akseer, Deena H AlAsfoor,
Ziyad Al-Aly, Fares Alahdab, Khurshid Alam, Ammar Albujeer,
Kefyalew Addis Alene, Raghib Ali, Syed Danish Ali, MehranAlijanzadeh,
Syed Mohamed Aljunid, Ala’a Alkerwi, Peter Allebeck,
Nelson Alvis-Guzman, Azmeraw T Amare, Leopold N Aminde,
Walid Ammar, Yaw Ampem Amoako, Gianna Gayle Herrera Amul,
我們的結果表明需要重新考慮酒精控制政策和健康計劃，並考慮放棄酒精的建議。減少人口的酒精使用，世界衛生組織提供了一套最好的購買政策，提供健康的一年飲酒費用低於一般人的平均成本收入。各國政府應該考慮這些問題建議可以在當地實施更廣泛的政策平臺，包括消費稅對酒精徵稅，控制酒精的物理可用性，酒精和銷售時間，以及控制酒精廣告。任何這些政策行動都會有所貢獻降低人口消費水平 - 減少與酒精使用有關健康損失。未能解決酒精使用的危害，尤其是在高消費水平下，會產對人口健康生可怕的影響。俄羅斯的死亡危機是一個引人注目的例子，酒精使用是主要的從20世紀80年代開始，是死亡人數增加的罪魁禍首，導致15-55歲男性死亡人數的75％。目前的全球趨勢 - 即人口老齡化 - 預示著酒精可歸因負擔不斷增加。在沒有政策的情況下，特別是許多癌症不成比例地影響老年人。所以，中低端SDI國家可以從政策中受益，今天的行動是保持酒精消耗低防止將來更大的健康損失。高和高中級SDI位置需要考慮更強減少酒精的政策，由世衛組織編寫，旨在減少人口消費水平。我們的結果應該在上下文中解釋該研究的侷限性。首先，我們的消費估計可能無法完全捕獲非法生產或未記錄，這是考慮到我們使用銷售數據進行估算。我們一直在努力調整消費以外的銷售額數據，但鑑於這些有估計的異質性，它可能是其他方法錄消費的量化。第二，一年內飲用模式被認為是一致的;然而，過去的工作表明飲酒模式，而不是平均水平，可能是標準日常消費這將與不同程度的風險和傷害有關。不幸的是，評估此類飲酒的數據要求年齡，性別和地點的模式遠遠超過了目前可用方法。例如，很少有前瞻性研究，量化飲酒模式和平均水平的影響消費水平串聯，要求正確評估酒精歸因的風險結。第三，用於估算機動車輛的數據僅限酒精使用給他人造成的傷害使用於美國（即FARS數據）。雖然是不太可能在FARS中觀察到的模式是巨大的與其他地方不同（附錄1，pp 141-144），這個假設需要進行測試，到更具體地點的估算。第四，我們無法做到找到來自酒精的人際暴力關於對他人造成的傷害的可靠數據，這是健康損失的潛在來源。更多的回顧性研究需要評估通過他人造成的傷害個人的酒精使用。第五，消費未評估年齡小於15歲的人群，因為這些年齡段的酒精使用數據稀少。在缺乏此類數據的情況下，可能採取的方法解決這個限制，例如假設消費老年人群體的模式或試圖推斷過去酒精消費水平很可能會引入額外的偏見或錯誤。更多青少年飲酒研究估計酒精需風險 - 該年齡組的可歸因負擔。最後，我們尋求量化酒精使用的風險僅用於結果符合比較風險標準的證據GBD研究的評估方法。但是，有其他結果，其中積累的證據表明酒精使用可能是一個風險因素。 31-33這些組合限制表明我們的結果很可能是 - 估計健康風險和總體歸因。
結論
酒精使用是疾病負擔的主要風險因素在全球範圍內，佔全球死亡人數的近10％在15-49歲的人口中，並且構成可怕的對未來人口健康的影響。今天的政策行動廣泛健康觀酒精的好處需要修改，尤其是改進的方法和分析繼續顯示大量的酒精使用導致全球死亡和失能。我們的結果表明，最安全的水平喝酒是沒有的。這個級別與大多數人都有衝突。其中包含健康益處，每天最多消耗兩杯。
———— / END / ————
入酒友群加微信shoucheng234
往期精彩回顧
中國白酒：能喝出健康來嗎？
中國白酒中的有毒物質
釀酒的奇蹟——極臭而後至極香
閱讀更多 休閒讀品雜誌社 的文章
