La Asociación Americana del Corazón acaba de publicar hoy sus recomendaciones sobre el alcohol, que coinciden con el informe de las Academias Nacionales.
El consumo moderado de alcohol, hasta 7 bebidas a la semana (1 al día), no entraña riesgos
Alcohol Use and Cardiovascular Disease: A Scientific Statement From the American Heart Association
TRENDS IN ALCOHOL USE
Alcohol
consumption remains extremely common in the United States. Eighty-five
percent of US adults report consuming alcohol at some point in their
lives,1 with per capita alcohol consumption estimated at 2.5 gallons/y.2,3
A nationally representative study suggests that the prevalence of any
alcohol use, including heavy alcohol use (≥5/4 [men/women] drinks on any
given day or ≥15/8 [men/women] drinks/wk), increased during the
COVID-19 pandemic (2018 versus 2020). The absolute increase for any
alcohol use was 2.7% (relative increase, 4.0%) and for heavy alcohol use
was 1.0% (relative increase, 20%), with these increases sustained in
2022.4
Initiation of alcohol use rapidly escalates through the teenage years, and prevalence rates peak in the early 20s.1 Recent (past-month) drinking is reported by >50% of US adults >18 years of age and 44% of those >65 years of age,1
indicating that alcohol use often persists across multiple decades of
life. In addition, 61.4 million people ≥12 years of age (22% in this age
group) reported binge drinking in the past month, defined as ≥5
alcoholic drinks on the same occasion for male individuals or 4 or more
alcoholic drinks on the same occasion for female individuals. Although
most alcohol users do not develop alcohol use disorders (≥1 in 101), annual deaths potentially related to alcohol top 175 000 in the United States5; thus, risks from alcohol use extend far beyond alcohol use disorder.2
DEFINITIONS OF STANDARD DRINKS AND PATTERNS AND CURRENT GUIDELINES FOR DRINKING
In
the United States, a standard drink includes about 0.6 oz (14 g) of
alcohol. Globally, however, the definitions of a standard drink differ,
ranging from 8 to 20 g. Alcohol by volume varies among beverage types;
therefore, a standard drink equivalent differs by beverage type (Figure 1).
There is general agreement about harmful levels of drinking (Table 1),
but mixed messages exist concerning health risks associated with lower
levels of alcohol use such as 1 to 2 standard drinks/d (Figure 1).6–12 The World Health Organization has stated that “no level of alcohol consumption is safe for health.”7
Canada’s guidance on alcohol and health has somewhat followed suit,
endorsing “drinking less is better” and delineating risks associated
with different weekly levels of alcohol use.6
Meanwhile, the 2020 to 2025 Dietary Guidelines for Americans (US
Department of Agriculture) reflect ongoing uncertainty and equipoise
derived from available evidence and refrain from definitive claims on
health among low-level consumers.12
A meta-analysis of 36 RCTs
evaluating different levels of alcohol use in short- to medium-term
crossover and parallel-group designs (2865 participants with and without
hypertension) showed that consumption of <2 drinks/d compared with
no alcohol use was not associated with a significant change in BP.38
However, drinking ≥3 drinks/d resulted in significantly higher
(increasing with higher alcohol consumption) SBP and diastolic BP (DBP)
compared with no or less drinking. In participants who consumed ≥6
drinks/d and reduced their intake on average by ≈50%, strong reductions
in SBP and DBP were observed (mean SBP and DBP differences, −5.5 mm Hg
[95% CI, −6.70 to −4.30] and −3.97 mm Hg [95% CI, −4.70 to −3.25],
respectively). These observational findings demonstrate the benefit of
abstinence or a significant reduction in alcohol intake on increased BP.
Because the number of women in such trials ranged from only 5 to 48 and
only 3 trials presented data specific to women, resulting in uncertain
pooled-effect estimates, a need to include more women in such research
clearly exists.38
A
recent meta-analysis of cohort studies found a dose-response
association between baseline levels of alcohol consumption and BP levels
after 5.3 years of follow-up, with no threshold regardless of sex and
geographic location.39
People who drank on average 1 drink/d exhibited a 1.25–mm Hg higher SBP
(95% CI, 0.5–2.0), whereas those who consumed a median of 3 drinks/d
experienced a 4.9–mm Hg (95% CI, 3.7–6.1) higher SBP.39 The increase in SBP was less for women compared with men.
In
a systematic review and dose-response meta-analysis of nonexperimental
cohort studies (>600 000 participants), there was a linear positive
association between alcohol intake and incidence of new-onset
hypertension above an alcohol intake of 1 drink/d (12 g/d).40
In most studies, hypertension was defined as a SBP/DBP ≥140/90 mm Hg,
and the follow-up period ranged from 2 to 22 years. Men demonstrated a
linear positive association between alcohol consumption and the risk of
hypertension compared with no alcohol intake, but the risk flattened at
intakes of >3 to 4 drinks/d.40
In women, however, an increase in the risk of hypertension was found at
an alcohol intake exceeding 1 drink/d, with a steeper slope at
increasing levels of alcohol consumption.40
ALCOHOL USE AND CORONARY ARTERY DISEASE
Many
studies have examined the relationship between alcohol use and the risk
of different subtypes of ischemic heart disease, including MI. The
effects of alcohol consumption on the risk and incidence of ischemic
heart disease (hereafter referred to as coronary artery disease [CAD])
are complex and vary depending on the amount and pattern of alcohol
intake and study design. In addition, how CAD was defined or coded (ie,
health records such as using International Classification of Diseases codes versus self-report) and which CAD outcomes were evaluated, for example, MI alone or grouped (eg, International Classification of Diseases, 10th Revision
codes that include angina, MI, complications after an MI), may
demonstrate different risk relationships with alcohol use. For example, a
meta-analysis using individual participant data from 83 observational
studies found that compared with the reference group (those consuming
between 0 and 25 g/wk, ≈1.78 drinks/wk), any consumption above this
level (up to 21 drinks/wk) was associated with a lower risk of MI
incidence.33 However, in evaluations of CAD that excluded MI, there was a positive and linear association with ≥7 drinks/wk.33
One of the earliest meta-analyses conducted to examine a dose-response
relationship reported that alcohol consumption (≤1 drink/d for women, ≤2
drinks/d for men) compared with no alcohol consumption was associated
with a lower (14%–25%) risk of incident CAD.41
In a recent MR study using traditional (linear) and nonlinear MR
techniques and using a single-nucleotide variant associated with alcohol
use disorder (as a means to examine the association between a
genetically predicted level of alcohol consumption), Biddinger et al42
reported that levels of alcohol intake (3–6 drinks/wk) were associated
with no to minimal increases in CAD risk but that risk began to rise at
levels exceeding 7 drinks/wk.
In a recent burden-of-proof approach, Carr et al43
examined the relationship between alcohol consumption and CAD risk from
case-control, cohort, and MR studies. This type of analysis
incorporates a 6-step framework for conducting a meta-analysis and
allows the generation of conservative estimates and interpretations of
risk-outcome relationships.44
The main findings from this analysis indicated that data pooled from
case-control studies (n=27) were associated with a 13% average reduction
in CAD risk across average consumption levels of 0 to 3 drinks/d,
whereas the risk reduction found from cohort studies (n=95) was lower
(5%) across average consumption levels of 0 to 3.5 drinks/d. For case
and cohort studies reporting only MI and analyzed together, the lower
relative risk across 0 to 50 g/d was ≈0.40 to 0.80. This analysis
revealed no statistically significant associations within the included
MR studies.43
These authors concluded that “data pooled from cohort and case-control
studies showed weak association between average levels of alcohol
consumption (≤50 g/day) and reduced CAD risk relative to no alcohol
intake, while data pooled from MR studies showed no association between
genetically predicted alcohol consumption and CAD risk.”43
Drinking
patterns, in particular binge and heavy episodic drinking, may modify
the relationship between alcohol consumption and CAD risk. For example,
moderate drinkers (<30 g/d) without a history of heavy drinking
occasions had a pooled relative risk for CAD incidence of 0.64 (95% CI,
0.53–0.71) compared with lifetime abstainers. However, for moderate
drinkers engaging in heavy episodic drinking occasions, the pooled
relative risk was 1.12 (95% CI, 0.91–1.37), demonstrating that consuming
<30 g/d was not associated with reduced CAD risk when accompanied by
occasional heavy or binge drinking.45
To
summarize, observational studies generally suggest lower risk of
incident MI among individuals who consume alcohol consumption within the
recommended US Dietary Guidelines for Americans limits compared with
lifetime abstainers, but methodological challenges in the interpretation
of these studies sharply limit the certainty of a causal relationship.9
A recently initiated clinical trial of abstention compared with
continued drinking among Spanish adults includes MI in its composite end
point; if the trial is successful in its recruitment, retention, and
adherence efforts, it will provide important evidence about the
potential causality of this relationship.46
ALCOHOL USE AND STROKE
A
meta-analysis of 27 prospective epidemiological studies found that
compared with no alcohol use or occasional drinking, low to moderate
drinking (≤2 drinks/d) was associated with a modestly (8%–10%) decreased
risk of ischemic stroke, whereas high (2–4 drinks/d) or heavy (>4
drinks/d) drinking was associated with an increased risk of all stroke
types.47 Similar findings were reported in an earlier meta-analysis48 and in later large-scale epidemiological studies.42,49–51
Analyses confined to current drinkers have revealed a positive
dose-response relationship between alcohol consumption and stroke risk,33,50 with consistent findings regardless of stroke types, for nonfatal and fatal stroke and for men and women.33
For example, in a combined analysis of 3 large-scale data sources in 19
high-income countries, a 100–g/wk (≈1 standard drink/d) increase in
alcohol consumption among ≈600 000 current drinkers increased the risk
of ischemic stroke by 13%, intracerebral hemorrhage by 17%, and
subarachnoid hemorrhage by 9%.33 The overall stroke risk increased by 15% in men and 9% in women per 100–g/wk increase in alcohol consumption.33
Furthermore, MR studies in populations of European and Asian ancestries
have reported a positive association between genetically predicted
alcohol consumption and risk of total stroke,42,52 ischemic stroke,50,52,53 intracerebral hemorrhage,50,52 and subarachnoid hemorrhage.52
Overall,
evidence from both observational and MR studies indicates that heavy
alcohol consumption (>4 drinks/d) is a risk factor for all stroke
types. However, the evidence is currently insufficient to draw
definitive conclusions about the relationship between moderate alcohol
consumption (≤2 drinks/d) and ischemic stroke.
ALCOHOL USE AND ARRHYTHMIAS
Atrial Fibrillation
Meta-analyses have concluded that heavier alcohol consumption predicts a heightened risk of developing atrial fibrillation (AF).54,55 Instrumental variable analyses have suggested causal relationships between alcohol exposure and AF in large populations.56–58
No clear threshold effect has been identified, nor has any one type of
alcoholic drink been differentially implicated; the relationship appears
to be fairly linear.55 Data remain conflicting as to whether 1 drink/d on average influences AF risk.59,60 Alcohol abstainers appear to be at a lower risk of AF than those who continue to drink,61
and a prospective, randomized trial of Australians consuming at least
10 drinks/wk demonstrated a substantial reduction in AF burden among
those instructed to abstain.62
Patients with paroxysmal AF fitted with continuous alcohol sensors
demonstrated a heightened odds of AF occurring within hours of a
drinking event,63
and a per-protocol analysis of randomized N-of-1 trials of AF triggers
found alcohol to be the only prespecified trigger associated with AF.30
Alcohol
consumption has been associated with left atrial enlargement and
fibrosis, both potential mediators of the alcohol-AF relationship.64–66 More often, those with alcohol-related AF may also have vagally mediated episodes.67
A randomized, double-blind, placebo-controlled study in humans
demonstrated that alcohol acutely shortened pulmonary vein refractory
periods.68
Supraventricular Tachycardia
Premature Ventricular Contractions, Ventricular Tachycardia, and Ventricular Fibrillation
Sudden Death
Long-term
consumption of ≈1 drink/d has been associated with the lowest risk of
sudden death, whereas heavy alcohol consumption may be associated with a
heightened risk of sudden death.49,72,74,75 Because sudden deaths are now known to have multiple causes,76 the underlying mechanisms may be related to alcoholic cardiomyopathies, MI, or other noncardiac causes of death.77
Bradycardias
Longitudinal cohort studies have not found a relationship between alcohol consumption and a higher risk of sinus node disease.78,79 However, moderate alcohol consumption has been associated with a lower risk of developing sinus node disease.78
Longitudinal cohort studies also suggest that alcohol either has no
effect or has a protective effect on atrioventricular conduction
disease.69,78,80,81
ALCOHOL USE, CARDIOMYOPATHY, AND HF
Long-term
excessive alcohol use is associated with the development of a dilated
left ventricle, normal or reduced left ventricular wall thickness and
mass, and, in advanced stages, HF with reduced left ventricular ejection
fraction.16
Current cardiomyopathy schemas recognize excessive alcohol use as a
nongenetic cause of dilated cardiomyopathy but also note that alcohol
may act as an epigenetic trigger in the presence of underlying genetic
variants.82
In the United States, prevalence estimates are elusive and vary among
reports. Data from the National Inpatient Sample show that among 352
million estimated hospitalizations of adults, 68 per 100 000 were
associated with the diagnosis of alcoholic cardiomyopathy (ACM). This
prevalence was higher among men than women.83
The average age of hospitalization is similar for men and women (≈56
years); however, when stratified by race, both Hispanic men and women
were hospitalized at a younger mean age (53 years).
The
exact amount and duration of alcohol consumption associated with the
development of ACM remain unknown. Data derived chiefly from
case-control studies suggest that consuming ≈7 to 15 standard drinks/d
over a 5- to 15-year period is associated with adverse changes in
systolic or diastolic ventricular function.84 Recently, however, in an observational cross-sectional study, Daka et al85 reported that as few as 4 drinks/wk (in the past month) was associated with an increased odds of diastolic dysfunction.
The
risk of ACM among those who consume alcohol is highly heterogeneous,
suggesting important interactions with other environmental exposures or
genetic factors. The presence of truncating variants in the gene
encoding the giant sarcomeric protein titin (the most common cause of a
dilated cardiomyopathy), may represent a genetic predisposition and
increased vulnerability to ACM, particularly in individuals reporting a
history of alcohol intake at ≈6 drinks/d over a 5-year period.86 Women appear to be at risk of ACM when exposed to lower amounts and shorter durations of alcohol use compared with men.87
Controversy
exists on the relationship between alcohol use and the risk of
developing HF. Several meta-analyses and prospective cohort studies
reported that <1 and 1 to 2 drinks/d were inversely associated with
incident HF.88–90
In contrast, an MR study reported both the absence of an association
and an inferred increased risk of HF attributed to alcohol.42
In the same MR study, when abstainers were excluded, exceeding 7 to 10
drinks/wk was associated with an increased risk of incident HF.42
Similarly, studies of individual participant data from large-scale
sources and 19 high-income countries revealed that usual alcohol
consumption of <7 drinks/wk was not associated with increased HF
risk, whereas levels of ≈21 drinks/wk were associated with an ≈50%
increase in HF risk.33
Among
studies and different populations, there appears to be a consensus that
exceeding 21 drinks/wk is associated with an increased risk of HF.
Recently, and similar to others, Wong and colleagues91
reported that there was no protective or adverse association between
alcohol consumption (≤5 drinks/wk) and the risk of developing HF.
However, in patients with structural or functional cardiac
abnormalities, alcohol use (≥5 drinks/wk) was associated with an
increased risk of progression of asymptomatic left ventricular
dysfunction and symptomatic HF (odds ratio, 5.0 [95% CI, 1.7–15.5]) over
a mean follow-up of 5.4 years.91
ALCOHOL USE AND SPECIAL POPULATIONS
Some
populations may be differentially affected by alcohol consumption
related to their sex (women), age (older adults), predominant pattern of
drinking (binge drinking), or presence of chronic diseases such as
diabetes. For example, women appear to be at risk of ACM when exposed to
lower amounts and shorter durations of alcohol use compared with men.87
The reasons for these sex differences remain unresolved but as noted
may be attributable to differences in the absorption, volume of
distribution, and first-pass metabolism of alcohol between men and
women, which give rise to differences in blood alcohol concentration
after the consumption of similar amounts of alcohol. Populations of
African and Asian ancestry may also experience different effects due to
genetic polymorphisms in alcohol-metabolizing proteins that may increase
the production of acetaldehyde, an aversive and mutagenic intermediate
metabolic product.92
Although studies are limited, follow-up of the National Health
Interview Study suggested that the lowest mortality risk occurred at
lower levels of consumption among Black individuals than White
individuals.93
Older
adults (50–80 years of age) have been included in many of the studies
reviewed here; however, other factors such as health status and
prescription drug use may affect the effects of alcohol on
cardiovascular conditions. Older adults may be vulnerable to
alcohol-medication interactions because of changes in metabolism and
increased likelihood of taking multiple medications. This is important
because among adults 50 to 80 years of age, 67% reported drinking
occasionally in the past year.94 Among those who drank, 27% (1 in 4) reported consuming ≥6 drinks on at least 1 occasion in the past year.94
Numerous
reports indicate that young adults (18–30 years of age) engaging in
binge drinking exhibit early signs of CVD such as endothelial
dysfunction (a precursor to atherosclerosis),95 coronary calcification,96 and arterial stiffness.97 Others have reported that binge drinking in young adults was associated with higher SBP,98,99
and binge drinking in adolescence (12–18 years of age) was associated
with a higher odds of high SP (SBP ≥130 mm Hg /DBP ≥80 mm Hg) in young
adulthood (24–32 years of age).100
In addition, several case-control studies have reported that binge
drinking was associated with an increased risk of different stroke types
in young adults.46,101,102
In an analysis of INTERSTROKE data from 32 countries, binge drinking
was associated with an odds ratio of 5.44 (95% CI, 1.81–16.4) of
ischemic stroke among patients ≤45 years of age.102 The potential for arrythmias among young adult binge drinkers was examined in the prospective MunichBREW II study.103
Over a voluntary binge drinking episode (blood alcohol levels ≥120
mg/dL), young adults (N=202; mean age, 29.9 years) were monitored with a
48-hour electrocardiogram encompassing the time frame during and after a
binge episode. The main findings were a significant increase in heart
rate (maximum rate, 97±16 bpm 4 hours after consumption) and a
significant increase in the percentage of atrial tachycardia beats. Only
1 participant developed AF, and 4 other participants experienced
clinically relevant supraventricular and ventricular arrhythmias.103
On the basis of these findings, there is a low occurrence of “holiday
heart syndrome,” a term used to refer to arrhythmias associated with
binge-like drinking episodes in young adults.
There
is a high rate of the co-occurrence of diabetes and CVD. In patients
with diabetes, a few RCTs have examined the effects of alcohol intake on
glycemic control and other cardiovascular parameters in patients.
Cardiovascular Diabetes and Ethanol randomized patients with type 2
diabetes (N=22; age, 40–75 years) to a low-risk drinking group (1
drink/wk of either red or white wine) or mineral water with a
Mediterranean diet/dinner for 2 years.104
Findings indicated that this level of alcohol consumption was safe, and
there were some small improvements in lipoproteins in the red wine
group (a 2.0–mg/dL increase in high-density lipoprotein cholesterol) and
a decrease in fasting glucose (−17.2 mg/dL) in the white wine group; no
changes were found among groups in BP or measures of adiposity.104
A crossover study reported that 4 weeks of red wine consumption (≈2
drinks/d) in men (n=19) and women (n=5) with type 2 diabetes
significantly increased awake SBP (2.5±1.2 mm Hg) but had no effect on
glycemic control, high-density lipoprotein cholesterol, fibrinogen,
C-reactive protein, or plasma homocysteine.105
In a 2-year trial among adults with type 2 diabetes, no significant
progression was observed in carotid plaque volume among individuals
randomized to water, white wine, or red wine (P=0.9 between groups).106
In a post hoc analysis among individuals in the highest tertile of
baseline carotid plaque volume, wine (red and white) significantly
reduced plaque volume over 2 years compared with baseline, whereas no
changes were found in the water group.106
PHYSIOLOGICAL MECHANISMS AND OTHER MEDIATING FACTORS ASSOCIATED WITH ALCOHOL CONSUMPTION
The
physiological mechanisms and other mediating factors associated with
alcohol use and cardiovascular outcomes are complex and multifaceted and
depend on the quantity and pattern of consumption. Some key mechanisms
related to CVD may include effects on lipid profiles, influence on
coagulation and thrombosis, and metabolic (ie, fasting glucose and
insulin) alterations, as well as body weight. Other mediating factors
might include socioeconomic status (SES), drinking with meals, and
presence of other health factors.
Table 2 notes the directional (significant) changes in selected cardiometabolic parameters summarized from 1 meta-analysis107 and another study using data from 3 cohorts of US adults.108 Glycemic data were derived from a systematic review of moderate alcohol consumption (in most studies, defined as 1–2 drinks/d).109 In addition, alcohol consumption (2 drinks/d) was found to reduce collagen-stimulated platelet aggregation110 while heavier drinking (1.5 g/kg, ~ 7 drinks) was associated with acutely increased platelet aggregation.111 Simultaneous intake of aspirin (325 mg) with alcohol (3–4 drinks) potentiates aspirin-induced bleeding time.112
https://www.ahajournals.org/doi/10.1161/CIR.0000000000001341
| Biomarker | Change reported in study |
|---|---|
| Lipid profiles | |
| Total cholesterol | No significant change107,108 |
| High-density lipoprotein | ↑ 3 mg/dL107 ↑ 7%108 |
| Low-density lipoprotein | ↓ −3.09 mg/dL107 No significant change108 |
| Triglycerides | No significant change107 ↓ −2.1%108 |
| Lipoprotein(a) | ↓ −0.04 g/L107 |
| Apolipoprotein A1 | ↑ 0.04 g/L107 |
| Hemostatic factors | |
| Plasminogen activator inhibitor 1 | ↑ 4.33 ng/mL |
| Fibrinogen | ↓ −0.13 g/L107 |
| E-selectin | No significant change107 |
| Inflammation | |
| C-reactive protein | No significant change107,108 |
| Interleukin 6 | ↓ −0.43 pg/mL107 |
| Tumor necrosis factor-α | No significant change107 |
| Adipokines | |
| Leptin | No significant change107 |
| Adiponectin | ↑ 0.06 mg/L107 ↑ 4.3%108 |
| Glycemic parameters | |
| Fasting glucose | No significant change109 |
| Fasting insulin | No significant change109 |
| Hemoglobin A1c | No significant change109 |
| Estimated insulin sensitivity | No significant change109 |
*
Data were extracted from 1 meta-analysis107
and another study using data from 3 prospective cohorts of US adults
(Nurses’ Health Study, Nurses’ Health Study II, and Health Professionals
Follow-up Study).108 Huang et al107
conducted a meta-analysis that included controlled short-term
intervention studies (1–8 weeks) comparing moderate alcohol consumption
(1–2 drinks/d) with no alcohol use. Values are median change values in
designated biomarker.107 Studies that included heavy drinking were excluded.107 Li et al108
compared individuals with moderate alcohol intake (average alcohol
intake, 3.3 servings/wk) with nondrinkers or those who drank less than 1
serving/d. Values are percentage differences in log-transformed values.108 Glycemic data are from Schrieks et al.109
Obesity
is an important cardiovascular risk factor; several studies have
examined the relationship between obesity and alcohol use, and findings
are equivocal.113
The most recent meta-analysis to date included 127 studies, and
findings differed when cohort and cross-sectional studies were analyzed
separately. Among the cohort studies analyzed, no association with
alcohol intake was found for risk of overweight (body mass index ≥25–30
kg/m2), obesity (body mass index ≥30 kg/m2), and overweight/obesity (body mass index ≥25 kg/m2).
However, among the cross-sectional studies analyzed, alcohol intake was
found to increase the odds for being overweight and overweight/obese,
but no association was found for the odds of being obese. In this same
study and in a dose-response analysis among cross-sectional studies,
light alcohol use (<14 g/d) had no association with being overweight,
obese, and overweight/obese. Similarly, alcohol use (14–28 g/d) had no
association with being overweight or overweight/obese; however, there
was 16% lower odds of being obese. Heavy (>28 g/d) alcohol use was
positively associated with being overweight and overweight/obese, but no
association was found with obesity. As the authors of this study note,
heterogeneity among studies was high, and examining the relationship
between alcohol use and weight is complicated and mediated by many
factors such as physical activity, dietary habits, genetics, and sex.
Alcoholic beverages can be a large source of calories: 1 g of pure
alcohol (ethanol) contains 7 calories. A standard glass of wine on
average equals 158 calories, and some alcoholic beverages may contain
≥500 calories. Therefore, it seems that addressing problematic alcohol
use should be an important component of obesity prevention and
management strategies.
Findings from some
studies suggest that SES and drinking during meals may be important
mediating factors of alcohol use and cardiovascular risk association.
For example, Ortolá et al31
examined the association of different alcohol consumption levels with
12-year CVD mortality among older adults (≥60 years of age) with
consideration of SES factors, beverage preference, and drinking with
meals. Compared with occasional drinkers (≤2.8 g/d), low (≥20 g/wk–20
g/d for men and ≥20 g/wk–10 g/d for women) and moderate (≥20–40 g/d for
men and >10–20 g/d for women) drinking was not associated with CVD
mortality risk, whereas high-risk alcohol consumption (≥40 g/d for men
and ≥20 g/d for women) was associated with higher CVD mortality (hazard
ratio, 1.21 [95% CI, 1.04–1.41]). CVD mortality was higher in those with
lower SES status and the presence of health-related factors; however,
wine preference and drinking during meals modified and negated the
association of mean alcohol intake with CVD mortality in all drinkers
with lower SES and the presence of health-related factors but especially
in high-risk drinkers.
Last, the way in
which alcohol affects vascular, myocardial, and brain function may
include many changes in cellular and molecular mechanisms. However,
reviewing these mechanisms is beyond the scope of this scientific
statement, but it would be interesting to explore them in future work.
Several reviews have been published that include details related to
cellular and molecular mechanisms potentially involved in mediating the
effects of alcohol.16,114
PATIENT EDUCATION: WHAT DO WE TELL THE PATIENT?
Providing
specific alcohol use recommendations for individuals with and without
CVD conditions remains challenging because of the lack of data from
well-conducted RCTs. If patients are currently consuming alcoholic
beverages, Table 3 summarizes evidence that could be shared with patients.115,116
In addition, it could be communicated to patients that consuming
alcoholic beverages with meals will slow down the absorption of alcohol,
as well as alternating alcoholic beverages with nonalcoholic beverages.
| BP | Alcohol consumption, even in moderation (1–2 drinks/d), can exacerbate high BP, and for some individuals, reducing or eliminating alcohol intake may be an important part of managing hypertension. |
| Stroke | Heavy (≥5/4 [men/women] drinks on any given day or ≥15/8 [men/women] drinks/wk) alcohol consumption increases the risk of all types of stroke. The impact of ≤2 drinks/d on stroke risk remains uncertain because of conflicting results. The association may partially be mediated by increased BP.33 |
| CAD | Consuming alcohol at or below the 2020 to 2025 Dietary Guidelines for Americans (≤2 drinks/d for men and ≤1 drink/d for women) may provide some risk reduction for CAD, and heavy or binge drinking is associated with an elevated CAD risk. |
| Arrhythmias | Data remain
conflicting as to whether 1 drink/d influences AF risk. However, alcohol
abstainers appear to be at a lower risk of AF than those who continue
to drink. A prospective trial of individuals with AF demonstrated a
substantial reduction in AF burden with abstinence. Long-term consumption of ≈1 drink/d has been associated with the lowest risk of sudden death, whereas heavy (≥5/4 [men/women] drinks on any given day or ≥15/8 [men/women] drink/wk) alcohol consumption may be associated with a heightened risk of sudden death. |
| ACM | Long-term (5–15 y) excessive (≥7–15 drinks/d) drinking can be associated with the development of ACM. However environmental and genetic factors may modify the development of ACM. In patients who received guideline-concordant HF therapy and reduced their alcohol consumption to <80 g/wk (≈6 drinks), there was an improvement in ventricular function.115,116 In addition, there was a better prognosis.115,116 |
| HF | Low to moderate (1–2 drinks/d) amounts of alcohol consumption do not increase the risk for HF, but heavy or binge drinking may increase the risk. In patients with structural or functional cardiac abnormalities, alcohol use (≥5 drinks/wk) was associated with an increased risk of progression of asymptomatic left ventricular dysfunction and symptomatic HF. |
| DCM | The safe and harmful levels of alcohol consumption remain unknown for patients with DCM, especially women.* |
ACM
indicates alcoholic cardiomyopathy; AF, atrial fibrillation; BP, blood
pressure; C
AD, coronary artery disease; DCM, dilated cardiomyopathy; and
HF, heart failure.
*
In patients with DCM, Tayal et al117
found that prior (baseline) moderate to excessive alcohol consumption
of (≈1–5 drinks/d for women and ≈2–5 drinks/d for men) was not
associated with adverse cardiovascular outcomes during the median
follow-up of 3.9 years (hazard ratio, 1.29 [95% CI, 0.73–2.26] for the
composite end point of cardiovascular mortality and HF and arrhythmic
events). However, among patients who met this composite end point
(n=78), only 15 had a history of moderate alcohol excess.117
Some
health care systems have begun to document alcohol use as a seventh
vital sign. Screening for alcohol use and discussing what constitutes
hazardous and harmful use (Table 1)
are relevant to all patients. It is important to note that among
patients surveyed in cardiology services (N=498), 66.3% were exposed to
reports that indicated moderate drinking can be good for the heart.118
It is important to deliver the right message because that study also
reported that exposure to reports of healthy-heart effects or mixed
messages about the cardiovascular effects of alcohol were associated
with increased odds of hazardous alcohol use (odds ratio, 1.67 [95% CI,
1.02–2.74]).118
Clinicians should reinforce healthy lifestyle behaviors such as
regularly engaging in physical activity, avoiding tobacco use, and
maintaining healthy body weight to achieve cardiovascular health and to
prevent disease. Other considerations for consuming alcoholic beverages
also can be shared with patients (Figure 1).
CONCLUSION, KNOWLEDGE GAPS, AND FUTURE DIRECTIONS
The
overwhelming evidence demonstrates that heavy (generally >2
drinks/d) and binge alcohol consumption is harmful to cardiovascular
health. Abstinence and reducing alcohol intake may lower the risk for
certain cardiovascular conditions such as hypertension. Uncertainty
remains about the true cardiovascular risk of drinking lightly such as 1
to 2 drinks/d (Figure 2
provides a graphical summary of data). Multiple unknowns persist,
providing opportunities for important research into the effects of this
commonly consumed substance. Although this document cannot review every
knowledge gap, several are worth highlighting. More research into the
following areas is needed: the heterogeneous effects within individuals,
potentially driven by differences in SES status, other demographics,
interactions with other environmental or dietary exposures, concomitant
comorbidities, balancing overall health effects (eg, differential
effects related to CVD versus cancer), and genetics, as well as
effective strategies for behavior change to influence the healthiest
patterns of alcohol consumption, for which evidence in populations
without alcohol use disorders is especially lacking. In addition, to
help establish biological plausibility, more research is needed on the
cellular and molecular mechanisms that may underpin different patterns
and amounts of alcohol consumption (more research gaps are noted in Table 4).
Last, research using what has been established as the gold standard to
provide the highest level of evidence and RCT designs is needed to
better elucidate the true health effects of light to moderate alcohol
consumption (1–2 drinks/d).
A Ground Truth on Alcohol Intake
My review of the 3 major reports and podcast with Dr. Vivek Murthy
Beyond concordance that high consumption of alcohol poses health risks, and gender specificity (difference in amount of alcohol intake and relationship to risk for men vs women) the only main point of agreement for risk of alcohol intake and outcomes was for breast cancer in women, as seen in the graph above. Even with that you can see there was not agreement for the level of alcohol vs the magnitude of risk. For the National Academies report, the risk was only seen beginning with 1 drink/day among women. The Surgeon General and ICCPUD reports were concordant except for the magnitude of increased risk
