Transcript July-Aug 2004

Update on
Alcohol and Health
Alcohol and Health: Current Evidence
July–August 2004
www.alcoholandhealth.org
1
Studies on
Alcohol and
Health Conditions
www.alcoholandhealth.org
2
Alcohol, Hypertension, and
Cardiovascular Disease
Mortality
Malinski MK, et al. Arch Intern Med. 2004;164(6):623–628.
www.alcoholandhealth.org
3
Objectives/Methods


To assess the relationship between drinking and
mortality among patients with hypertension
Self-reported data from the Physicians’ Health Study


14,125 males with a history of current or past treatment
for hypertension and without myocardial infarction, stroke,
cancer, or liver disease at baseline
analyses adjusted for potential confounders (e.g., age,
smoking, diabetes)
www.alcoholandhealth.org
4
Results
Amount Consumed
Relative Risk (RR)*
of Cardiovascular
Disease Death
RR of Total
Death
1–6 drinks per week
0.6
0.7
≥1 drinks per day
0.6
0.7
*Compared with Those who Rarely or Never Drank
www.alcoholandhealth.org
5
Conclusions/Comments



In this cohort of physicians with hypertension
who consumed low amounts, moderate alcohol
consumption reduced risk of CVD mortality and
total mortality.
There are limits to applying these results to
broader primary care populations.
Clinicians should individualize recommendations
about alcohol use and help patients determine
whether the cardiovascular benefits from
moderate drinking outweigh any potential risks.
www.alcoholandhealth.org
6
Changes in
Alcohol Consumption
and Their Effects on
Mortality, Cancer, and
Coronary Heart Disease
Grønbaek M, et al. Epidemiology. 2004;15(2):222–228.
www.alcoholandhealth.org
7
Objectives/Methods


To examine how changes in intake affect
risk of mortality, coronary heart disease
(CHD), and cancer
Data from a longitudinal study



6644 men and 8010 women
completed 2 health exams at 5-year intervals
analyses adjusted for possible confounders
(e.g., age, sex, smoking)
www.alcoholandhealth.org
8
Results
Stable
drinking
Risk of All-Cause
Mortality
Risk of CHD Mortality
Risk of
Cancer
Mortality
U-shaped, with non-drinkers
and heavy drinkers having
the highest mortality
No significant association
with amount consumed
Significantly
higher in
heavy drinkers
Borderline significant risk in
non-drinkers
Increases
or
decreases
in intake
Significantly higher when
intake decreased to <1 drink
per week
No significant association
with increases
No significant
association
Borderline significant risk for
light drinkers who decreased
intake to <1 drink per week
Non-drinkers consumed <1 drink per week.
Light drinkers consumed 1-6 drinks per week.
Heavy drinkers consumed >13 drinks per week.
www.alcoholandhealth.org
9
Conclusions/Comments


This study found that moderate drinkers who
decreased their consumption to little or no
alcohol increased their risk of death from all
causes.
According to these findings, adult moderate
drinkers—to lower their risk of death—should
not be advised to reduce their intake or to
abstain.
www.alcoholandhealth.org
10
Drinking Patterns—
Not Just Total Amount—
Affect Risk
Trevisan M, et al. Addiction. 2004;99(3):313–322.
Tolstrup JS, et al. Addiction. 2004;99(3):323–330.
www.alcoholandhealth.org
11
Objectives/Methods
Trevisan study:

To determine the effect of drinking patterns on
the risk of myocardial infarction (MI)

Case-control study



427 white males with incident MI
905 healthy white male controls selected randomly
from 2 Western New York counties
analyses adjusted for multiple risk factors (e.g., age,
smoking, diet, physical activity)
www.alcoholandhealth.org
12
Results
Risk of MI among Current
Drinkers* compared with
Lifetime Abstainers
Drank <1 time during the week
OR 0.5
Drank daily
0.4
Drank on weekdays and
weekends (compared with only
on weekends)
0.5
Drank mainly with food
0.5
*past 1-2 years
www.alcoholandhealth.org
13
Objectives/Methods
Tolstrup study:


To learn how drinking patterns affect risk of allcause mortality
Questionnaire data from a population-based
Danish cohort


26,909 men and 29,626 women
analyses adjusted for multiple risk factors (e.g., age,
smoking, diet, physical activity)
www.alcoholandhealth.org
14
Results



Frequent drinkers (≥2 times per week) had a
lower hazard of death than non-frequent
drinkers.
Compared with males drinking <1 drink per
week, male frequent drinkers who consumed 1–
20 drinks per week had significantly lower
mortality.
Compared with females drinking <1 drink per
week, female frequent drinkers who consumed
larger amounts did not have significantly lower
mortality.
www.alcoholandhealth.org
15
Conclusions/Comments



These studies add to the accumulating evidence
that drinking patterns do matter.
Result of these and similar studies should be
viewed with caution because small numbers of
infrequent drinkers in the high consumption
categories make it difficult to adjust adequately
for total alcohol intake.
While they wait for better data, clinicians should
steer moderate drinkers towards more frequent
drinking of small amounts rather than less
frequent drinking of larger amounts.
www.alcoholandhealth.org
16
Risky Sex:
It’s the Alcohol
Ehrenstein V, et al. Drug Alcohol Depend. 2004;73(2)159–166.
www.alcoholandhealth.org
17
Objectives/Methods


To better understand alcohol’s effect on
inconsistent condom use
Data from a prospective cohort of adults with
HIV and current or past alcohol problems


345 subjects
analyses adjusted for possible confounders (e.g.,
race, education, other drug use)
www.alcoholandhealth.org
18
Results

Alcohol consumption was significantly associated
with inconsistent condom use (<100% over the
last 6 months).
% Reporting Inconsistent
Condom Use
Heavier drinkers (>14 drinks per
week for men, >7 for women)
51%
Moderate drinkers
41%
Abstainers
32%
www.alcoholandhealth.org
19
Results (cont.)
Factors associated with significant increases in risk of
inconsistent condom use (odds ratios between 2 and 4):



heavier drinking in users
of injection drugs
being female
identifying as gay or
lesbian



living with a partner
having 2 or more sexual
partners
agreeing that condoms
are a hassle to use
Factors associated with borderline significant increases in risk:
recently testing HIV-positive, selling sex for drugs or money,
having a higher CD4 cell count
www.alcoholandhealth.org
20
Conclusions/Comments



Whether the findings in this cohort (30%
employed, 67% non-white, 29% homeless) will
hold true for others with HIV is unknown.
Drinking >1–2 drinks per day appears to
increase risk of inconsistent condom use,
particularly among users of injection drugs.
When addressing this risk, clinicians, public
health practitioners, and others must consider
factors that contribute to inconsistent use—
many of which are modifiable.
www.alcoholandhealth.org
21
Lasting Effects of Alcohol
on Cognitive Function in
HIV Infection
Green JE, et al. Am J Psychiatry. 2004;161(2):249–254.
www.alcoholandhealth.org
22
Objectives/Methods


To better understand the effects of heavy
alcohol use on cognitive function in people
infected with HIV
Assessment of various neuropsychological
functions and symptoms of anxiety and
depression in 80 gay or bisexual men


50 with HIV (21 who also had past alcohol abuse or
dependence)
30 without HIV (12 who had past alcohol abuse or
dependence)
www.alcoholandhealth.org
23
Results


Overall cognitive impairment did not differ significantly
between those with and without HIV.
Men with past alcohol abuse or dependence, compared
with those without, showed cognitive impairment on a
significantly greater number of measures.
Mean # of Measures
indicating Cognitive
Impairment
Those with past abuse/dependence and HIV
5
Those with past abuse/dependence without HIV
3.8
Those without past abuse/dependence but with HIV
3.5
Those with neither past abuse/dependence nor HIV
2.9
www.alcoholandhealth.org
24
Conclusions/Comments



This study’s strength is its detailed assessments
of participants.
Given the study’s small sample size, researchers
could not adjust for factors that could have
explained the association between past alcohol
use disorders and cognitive impairment.
Nonetheless, the findings suggest that the
association between alcohol abuse or
dependence and cognitive dysfunction in HIVinfected men is particularly pronounced.
www.alcoholandhealth.org
25
Another Complication of
Drinking during
Pregnancy
Avaria MD, et al. J Pediatr. 2004;144(3)338–343.
www.alcoholandhealth.org
26
Objectives/Methods


To examine the effects of alcohol exposure
in utero on the peripheral nervous system
Study of 30 neonates born to women
identified during prenatal care

17 exposed neonates; 13 unexposed
www.alcoholandhealth.org
27
Results


No neonates had classic signs of fetal alcohol
syndrome; 1 had possible fetal alcohol effects.
Alcohol-exposed neonates had significantly slower
nerve conduction velocities and response
amplitudes in some nerves.


Differences were significant in the ulnar motor nerve and
borderline significant in the tibial motor nerve.
No significant differences were found in peroneal,
median, or sensory nerve conduction.
www.alcoholandhealth.org
28
Conclusions/Comments



This small study could not determine the length
and intensity at which alcohol exposure causes
harm, nor could it inform us of the clinical
implications of its findings.
Despite its sample size, the study did
demonstrate yet another complication of alcohol
consumption during pregnancy.
The best advice to pregnant women or to those
planning pregnancy is abstinence.
www.alcoholandhealth.org
29
Alcohol Consumption and
Diabetes Mellitus
Howard AA, et al. Ann Intern Med. 2004;140(3):211–219.
www.alcoholandhealth.org
30
Objectives/Methods


To summarize the findings of studies on
alcohol consumption and the risk and
control of diabetes
Systematic review

32 studies, 27 involving type 2 diabetes only
www.alcoholandhealth.org
31
Results



Moderate drinking (defined variably in each
study), compared with not drinking, reduced
diabetes risk in 11 of 18 studies.
Drinking (defined variably, but as little as 1.4
drinks per day) significantly increased diabetes
risk in 4 of 18 studies.
Consumption (about 1–6 drinks on an occasion
under experimental conditions) did not impair,
and sometimes improved, glycemic control.
www.alcoholandhealth.org
32
Results (cont.)



In patients with diabetes, moderate drinking, compared
with not drinking, significantly decreased risk of
 death from coronary heart disease (CHD) (by up to
79% in 3 of 4 studies);
 CHD-related events (by up to 55% in 2 of 3 studies).
Consuming about 3 drinks on an occasion with
troglitazone or a sulfonylurea did not affect glycemic
control.
Heavy use (defined variably) significantly increased
diabetic retinopathy in 1 study and had no effect in
another.
www.alcoholandhealth.org
33
Conclusions/Comments



Moderate alcohol use may decrease risk of type 2
diabetes and diabetes-related cardiovascular
events.
However, results from this review are inconsistent,
and some included studies suggest increased risk at
drinking levels often considered to be moderate.
Given current evidence, it is likely safe for people
with type 2 diabetes to follow usual drinking
recommendations (≤1 drink per day for women
and people over 65 years; ≤ 2 drinks per day for
men aged 65 and under).
www.alcoholandhealth.org
34
Studies on
Interventions
www.alcoholandhealth.org
35
Efficacy of
Brief Interventions in
Primary Care
to Reduce Risky and
Harmful Alcohol Use
Whitlock EP, et al. Ann Intern Med. 2004;140(7):557–568.
www.alcoholandhealth.org
36
Objectives/Methods


To inform clinical guidelines of the U.S.
Preventive Services Task Force (USPSTF) about
brief behavioral counseling interventions
Systematic review of studies on the efficacy of
these interventions

12 controlled trials included
www.alcoholandhealth.org
37
Results

Those receiving brief multi-contact interventions
(initial session up to 15 minutes and at least 1
follow-up) reduced their average weekly alcohol
intake by 13%–34% more than controls in 4 of 7
trials.


10%–19% more intervention participants than controls
drank safe amounts.
Very brief (up to 5 minutes) or brief single-contact
interventions were ineffective or less effective in
reducing risky or harmful alcohol use.
www.alcoholandhealth.org
38
Results (cont.)

All effective interventions included at least
2 of 3 key elements:


feedback, advice, and goal setting.
No significant differences were found
among men and women receiving brief
interventions.
www.alcoholandhealth.org
39
Conclusions/Comments


USPSTF gave a grade B recommendation (at least
fair evidence of improved health outcomes and
benefits outweighing potential harms) to
screening and brief counseling in primary care to
reduce risky or harmful alcohol use among adults.
Better strategies to implement screening and
more effective brief interventions in actual clinical
practice must be developed.
www.alcoholandhealth.org
40
Alcohol Screening and
Referral on
College Campuses
Foote J, et al. J Am Coll Health. 2004;52(4):149–157.
www.alcoholandhealth.org
41
Objectives/Methods


To assess alcohol screening and referral
on college campuses
Survey of a state-stratified random sample
of 4-year colleges/universities with health
centers

76% response rate; 234 schools analyzed
www.alcoholandhealth.org
42
Results



33% of the schools conducted routine alcohol
screening of most or all (96%) students visiting
their health centers.
Urban, large, and especially public schools were
significantly more likely to routinely screen.
Screening was typically part of a standard
medical history and physical.

12% of schools used a standardized instrument (most
often the CAGE) to screen.
www.alcoholandhealth.org
43
Results (cont.)

On average, health centers offered 3
referral options. The most common
included




the campus counseling center;
substance abuse treatment in the community;
12-step programs;
individual therapy.
www.alcoholandhealth.org
44
Conclusions/Comments


These findings highlight both the lack of alcohol
screening in college health centers as well as
potential inadequacies in practices used by those
who do screen.
Given the harm caused by—and few successful
strategies to curb—heavy drinking, college
health centers should receive the necessary
support to implement effective screening and
referral.
www.alcoholandhealth.org
45
Does Experience Influence
Hospitals’ Resource Use in
Treating Alcohol and
Other Drug Diagnoses
Bramble JD, et al. J Gen Intern Med. 2004;19(1):36–42.
www.alcoholandhealth.org
46
Objectives/Methods



To investigate whether a hospital’s experience
with alcohol and other drug diagnoses or its
teaching status impacts resource use
National data from 1996 Healthcare Costs and
Utilization Project and other large administrative
databases
Analyses adjusted for multiple hospital and
patient characteristics
www.alcoholandhealth.org
47
Results


Relative experience was significantly
associated with lower charges but also
associated (borderline significance) with
greater length of stay.
Teaching hospital status was not
significantly associated with either charges
or length of stay.
www.alcoholandhealth.org
48
Conclusions/Comments



The authors speculated that the lower charges
and longer lengths of stay in high volume
hospitals might have resulted from managing
patients with substance-related disorders in less
expensive settings.
They did not address the quality of care,
readmissions, or other outcomes.
Thus, whether this less expensive care leads to
better or worse clinical outcomes remains
unknown.
www.alcoholandhealth.org
49
Studies on
Special Populations
www.alcoholandhealth.org
50
Drinking Patterns and
Rates of Alcohol-Related
Problems in Three
Urban Populations
Bobak M, et al. J Epidemiol Community Health.
2004;58(3):238–242.
www.alcoholandhealth.org
51
Objectives/Methods


To examine whether drinking patterns, in
addition to overall consumption, contributed to
differences in alcohol-related problems
Interview data from 1118 men and 1125 women
randomly selected from Russian, Czech Republic,
and Polish population registries
www.alcoholandhealth.org
52
Results
Russian Men
Czech Men
Polish Men
Reported ≥2 positive
responses to the CAGE
35%
19%
14%
Reported ≥2 negative
consequences related to
consumption
18%
10%
8%
5 liters
9 liters
4 liters
Drinking sessions per year
67
179
79
Average consumption per
drinking session
mean 71 g
45 g
46 g
Mean annual intake
Russian men were also significantly more likely to drink ≥80, ≥120, or ≥160 g
of alcohol on an occasion.
www.alcoholandhealth.org
53
Conclusions/Comments


In this study, consuming large amounts per
drinking session explained a substantial part of
the differences in negative consequences among
the 3 populations.
These findings suggest that average
consumption alone does not determine alcoholrelated problems at the population level.
www.alcoholandhealth.org
54
Alcohol and Other
Drug Use Initiation in
American Indian
Adolescents
Novins DK, et al. J Am Acad Child Adolesc Psychiatry.
2004;43(3):316–324.
www.alcoholandhealth.org
55
Objectives/Methods


To examine the risk of initiation and
progression of substance use in American
Indian adolescents
Semiannual survey data collected over 3
years from 14 to 20 year olds at 7
predominantly American Indian high
schools
www.alcoholandhealth.org
56
Results

Of 568 adolescents who were abstinent at baseline,


42% initiated use with alcohol, marijuana, and/or
inhalants;
7% initiated use with other illicit drugs (e.g., cocaine).

Initiation of use peaked at approximately age 18.

Alcohol was the most common first substance used.
www.alcoholandhealth.org
57
Results (cont.)



Of 1244 adolescents who used alcohol,
marijuana, and/or inhalants at baseline, 24%
progressed to other illicit drug use.
Adolescents who had initiated substance use
with marijuana or inhalants (versus alcohol)
were significantly more likely to progress to
other illicit drug use.
Risk of initiation and progression significantly
differed by tribal group and season.
www.alcoholandhealth.org
58
Conclusions/Comments



All American Indian adolescents will not necessarily
follow the specified patterns of substance use initiation
and progression.
The findings of this study, however, do suggest that
almost half of American Indian adolescents initiate
substance use with alcohol, marijuana, and/or inhalants,
and a quarter of users progress to other illicit drugs.
Results imply that efforts to prevent initiation and
progression may need to be tailored to the specific
issues facing individual communities.
www.alcoholandhealth.org
59
Other
Studies
www.alcoholandhealth.org
60
Binge Drinking’s Effect
on Health-Related
Quality of Life
Okoro CA, et al. Am J Prev Med. 2004;26(3):230–233.
www.alcoholandhealth.org
61
Objectives/Methods


To examine the association between binge
drinking (consuming 5 drinks on 1 occasion in
the past month) and health-related quality of life
(HRQQL)
Data from the Behavioral Risk Factor
Surveillance System


cross-sectional, random-digit-dial telephone survey of
adults aged 18 and older nationwide
analyses adjusted for possible confounders
www.alcoholandhealth.org
62
Results

Frequent binge drinkers*,compared with non-binge
drinkers, were significantly more likely to
experience ≥14 unhealthy days in the past month.


Increased risk was mainly due to having more
mentally (vs. physically) unhealthy days.
Infrequent binge drinking** was significantly
associated with experiencing ≥14 unhealthy days
among women and people ≥55.
*had ≥3 episodes in the past month; **1–2 episodes in the past month
www.alcoholandhealth.org
63
Conclusions/Comments



Frequent, and sometimes infrequent, binge
drinking is associated with worse HRQOL and
mental distress.
It is not clear whether binge drinking led to
worse HRQOL or vice versa.
Findings underscore the adverse consequences
of binge drinking and support the need effective
prevention efforts.
www.alcoholandhealth.org
64
WHO Report Summarizes
Biological Factors Related to
Substance use
Geneva, Switzerland: World Health Organization, 2004
www.alcoholandhealth.org
65
Objectives/Methods

To summarize 30 years of research on
biological factors related to substance use
www.alcoholandhealth.org
66
Results



“…Substance dependence is a chronic, relapsing
disorder with a biological and genetic basis, and
is not simply due to lack of will or desire to quit.”
Tobacco, alcohol, and illicit drugs are responsible
for 9%, 3%, and 0.4% of deaths worldwide,
respectively.
Tobacco and alcohol are among the top 10
leading risk factors of avoidable disease burden,
with each responsible for approximately 4% of
disability worldwide.
www.alcoholandhealth.org
67
Results (cont.)



Effective treatments for substance dependence
are available and can be integrated into health
systems, including primary care.
Neuroscience-based knowledge of substance
dependence can reduce the strong and
pervasive stigma associated with substance use.
Proper treatment of either substance
dependence or mental illness requires attention
to the comorbidity of these conditions.
www.alcoholandhealth.org
68
Conclusions/Comments


This report summarizes what researchers
have known for some time—there is a
scientific basis for the etiology and
management of alcohol, tobacco, and
other drug use disorders.
Its findings may help dispel long-standing
myths about the addictions.
www.alcoholandhealth.org
69