Data from the national Youth Risk Behavior Survey indicates that most violence-related
behaviors among US high school students (e.g., physical fighting, carrying
a weapon on school property, physical fighting on school property) decreased
significantly between 1991 and 2003, corresponding with a decline in the
national youth homicide rate (CDC 2004). However, the prevalence of being
injured in a physical fight remained stable across subgroups during this
period, and the prevalence of being threatened or injured with a weapon
on school property increased among white and black students in recent
years (CDC, 2004). Moreover, preliminary data drawn from a new US surveillance
system, the national Violent Death Reporting System (NVDRS)—which
collects detailed information regarding violent deaths from multiple sources—indicates,
for the first six states participating in the NVDRS, substantial increases
in both homicide and suicide rates among males between 2000 and 2003 and
in homicide rates among young males (25 years or younger) between 2002
and 2003 (CDC, 2005).
POPULATIONS AT RISK
Specific populations that appear to be at elevated risk for injury attributable
to alcohol-fueled violence/aggression include couples in which at least
one partner drinks frequently, college students, young women, and bar
patrons.
Prior research has indicated that alcohol problems are common among partner violent
men, while men who drink heavily are more likely to perpetrate wife abuse
and women who abuse alcohol are more likely to be victims of violence.
Fifty to 60% of male alcoholics have been violent toward a female partner
in the year before alcoholism treatment (O’Farrell et al., 2004),
and at least two-thirds of women alcoholic patients have been the victims
of violence from a male partner in the previous year (Chase et al., 2003).
In data from a representative national sample of couples living in the
continental US, drinking problems—along with a history of being
physically harmed by parental figures during childhood and impulsivity—were
identified as risk factors for IPV in the US general household population
(Schafer et al., 2004). Data from a nationally representative US study
of alcohol consumption in young men indicate that alcohol use is strongly
related to marital aggression, with the presence of either low marital
satisfaction or high hostility sufficient, when combined with high scores
on a measure of alcohol dependence, to increase the risk for marital aggression
(Leonard and Blane, 1992). Moreover, a review of 60 studies that tested
the relationship between alcohol use and three marital functioning domains
(satisfaction, interaction, and violence), provided overwhelming support
for the notion
that alcohol use is maladaptive, and that it is associated
with dissatisfaction, negative marital interaction patterns, and high
levels of marital violence (Marshal, 2003).
Nearly 600,000 college students between the ages of 18 and 24 are assaulted by
another student who has been drinking (Hingson et al., 2005).
In the National College Women Sexual Victimization (NCWSV) study (Fisher et al.,
2000), utilizing a national telephone survey of randomly-selected women
attending a 2- or 4-year college or university in the U.S. during the
fall of 1996, nearly 5% of college women were found to have been sexually
victimized in any given calendar year, suggesting that a campus of 10,000
women might expect to see in excess of 350 rapes in a given academic year.
Research has also shown that women who regularly drink in bars experience
a substantial amount of bar-related physical and sexual aggression (Parks
and Zetes-Zanatta, 1999; Parks and Miller, 1997).
Researchers have found that city blocks that include bars have higher rates of
assaults, robberies, and rapes than do other city blocks, even after controlling
for unemployment and poverty (Roncek and Maier, 1991), and some observers
have concluded that alcohol outlet density may be related to violent assaults
when certain conditions prevail
(Gorman
et al., 1998; Treno et al., 2001; Speer et al., 1998; Scribner et al.,
1995, 1999; Alaniz et al., 1998).
RESEARCH PROGRESS
The Epidemiology of Alcohol Use and Violence
Approximately 40 percent of inmates incarcerated for murder report that they
had been using alcohol at the time of their offense (Greenfield et al.,
2001). As is the case with other violent crimes, alcohol use frequently
is found to precede homicides, but the potential causal relationships
that are involved are controversial, complex, and poorly understood and
studies have not conclusively demonstrated a causal role for alcohol availability
in the occurrence of violent events (NIAAA, 2000; Lipsey et al., 1997;
Parker and Rebhun, 1995; Parker and Cartmill, 1998). Clearly, more research
is needed in order to illuminate the ways in which alcohol availability
interacts with other factors in the causal process.
It appears that alcohol use is linked to violent behavior by way of interacting
pharmacological, physiological, endocrinological, genetic, situational,
psychological, environmental, and sociocultural determinants (Miczek et
al., 1993). Insofar as theoretical explanations that focus on only one
of these domains are likely to result in an incomplete understanding of
the role of alcohol in escalation to violence,
theoretical
models that bridge these frameworks are urgently needed (Chermack and
Giancola, 1997; Von der Pahlen, 2005; Graham et al., 2000).
Research on suicide has focused on five predisposing factors: aggression/impulsivity,
alcohol dependence, negative affect, interpersonal disruption, and depression
(Connor and Duberstein, 2004). Roughly the same constellation of factors
is discussed in the literature on suicidality among adolescents (Windle,
2004). However, treatments that address co-occurring alcohol dependence
and suicidality have been slow to develop (Esposito-Smythers and Spirito,
2004; Goldsmith et al., 2002). Biomedical research is just beginning to
explore two critical issues that bear on the connection between suicide
and alcohol dependence. COGA studies have suggested that chromosome 2
contains linkages for phenotypes of both alcohol dependence and suicidality
(Hesselbrock et al., 2004). Neuroscience research, however, indicates
that alcohol dependence and suicidality are related to separate serotonergic
and noradrenergic systems (Underwood et al., 2004).
Studies of behavior, cognition and expectancies
Psychological explanations of the means by which alcohol increases aggression
tend to focus on its effects on cognitive and attentional abilities, including
reactions to cues of punishment, decreased response flexibility, and reduced
intellectual functioning (Leonard, 2001; Bushman, 1997; Kantor and Jasinski,
1998; Steele and Josephs, 1990; Bushman and Cooper, 1990; Pihl et al.,
1984). The term “alcohol myopia” (Steele and Josephs, 1990)
has been used to describe alcohol-induced cognitive impairment, specifically
referring to the drinker’s inability to attend to inhibitory cues
and, therefore, more likely to act upon impelling cues which are thought
to require relatively little cognitive processing. Alcohol’s effects
on cognitive and attentional abilities are presumed to affect communication,
raising the likelihood of misinterpretation, perception of a threat, and
interpersonal violence as a defensive reaction (Ogle and Miller, 2004;
Pernanen, 1981, 1991; Sayette et al., 1993).
It is also clear that drinking of alcohol gives rise to certain expectations
in people that may increase the risk of aggression. When male and
female college students read vignettes depicting a young woman and man
consuming either alcoholic or nonalcoholic beverages (Abbey and Harnish,
1995), both genders of readers perceived the males and females in the
vignettes to be most sexual when they were drinking alcohol, suggesting
not only that alcohol consumption is seen as a sexual cue, but also suggesting
that alcohol use increases the odds that a woman’s expression of
platonic friendliness will be misperceived by a male companion as a sign
of sexual interest (Abbey et al., 2000). Similarly, in an experiment conducted
by Norris et al. (2002) with male social drinkers who were asked to read
a violent pornographic story, both the indirect effects of pre-existing
sex-related alcohol expectancies and the direct effects of expected and
actual alcohol consumption increased men’s self-reported likelihood
of behaving in a sexually aggressive manner. Furthermore, alcohol-related
cognitive impairments may make it difficult for a woman to correct any
misperceptions and to resist any unwanted sexual advances.
Availability
Any understanding of alcohol-related violence must recognize the primary importance
of alcohol availability. Findings from ecological studies have
demonstrated that homicides, along with other violent crimes, are higher
in areas where alcohol availability and outlet density are higher (Scribner
et al., 1999; Parker and Rebhun, 1995). However, this suggestive relationship
is obscured by a number of potential confounders, including urbanicity
and poverty. While research should clarify the role of availability relative
to these other factors, it should also capitalize on natural experiments
to investigate whether reductions in availability or outlet density are
subsequently followed by reductions in homicide rates. Cross sectional
studies have found high alcohol outlet density to be associated with violent
assaults, robberies and rapes (Scribner et al., 1995, Alaniz et al., 1998,
Stevenson et al. 1999). Further, Grube and Stewart (2004) identified six
prospective studies of changes in outlet density; however, results were
small and inconsistent. Most recently, Gruenewald and Remer (2006) reported
associations between alcohol-related violence and geographic outlet density
in a prospective study in California.
A systematic review of randomized controlled trials to reduce alcohol dependence
and abuse in the general population found reduced alcohol consumption
was accompanied by reductions in domestic violence and suicide attempts
(Dinh-Zarr et al., 1999). On the other hand, studies examining the effects
of increasing the legal drinking age increase on violence have been mixed.
While Engs and Hanson (1998) reported increases in fighting, Davis and
Reynolds (1990) found decreases in fighting, Joksch and Jones (1993) reported
reduction in disorderly conduct, vandalism and assaults, and Jones et
al. (1992) and Birkmayer and Hemenway (1999) noted reductions in suicides.
Reductions in violence have also been achieved by community-wide efforts to promote
more alcohol-safe environments. Holder et al. (2000) conducted interventions
in two California communities and one South Carolina community (each with
a matched control) to encourage responsible beverage service, enforce
drinking and driving laws, reduce underage access to alcohol, and limit
availability through local zoning powers. The results showed that self-reported
drinking-driving, night-time injury crashes, and assault injuries observed
in emergency departments all declined in the experimental communities.
Weitzman et al. (2004) studied ten colleges that implemented interventions
designed to affect alcohol availability, legal sanctions for alcohol-related
infractions, the quality of the physical settings where alcohol is consumed,
and the levels of alcohol advertising and promotion. Positive results
were obtained only on those campuses that more faithfully implemented
protocols to effect environmental change. Among those colleges there were
statistically significant reductions in self-reports of binge drinking,
vandalism, injuries, assaults, and unwanted sexual advances.
A landmark analysis examined the impact of local policy changes in Barrow, Alaska.
During a 33 month period, citizens of that community, through referendum
imposed, then withdrew, and then reimposed a total ban on alcohol sales.
There were significant decreases in emergency room visits (including those
due to assaults) during the period in which alcohol was banned. When the
ban was lifted, these incidents returned to the pre- ban level, and when
the ban was reimposed, these events declined again (Chui et al., 1997).
A number of studies suggest that levying excise taxes on alcohol can result in
significant reductions in rates of violence. Higher alcohol prices which
affect consumption levels have been found to reduce robberies and rapes
(Cook and Moore, 1993a, 1993b; Cook and Tauchen, 1982). Cook and Moore
(1993a, 1993b), Chaloupka et al. (1998) and Markowitz and Grossman (1998,
2000) all found that raising taxes on alcohol reduces problems (e.g. homicides,
suicides, rape, robbery, assaults, motor vehicle thefts, domestic violence
or child abuse). Markowitz and Grossman (2000) reported data suggesting
that a 10% increase in the excise tax on beer may reduce the probability
of severe child abuse by 2.1%.
RESEARCH QUESTIONS/ DIRECTIONS
Epidemiology of Alcohol Use and Violence
The goal for future research should be to move beyond establishing the presence
of alcohol prior to a crime and toward greater clarification of alcohol's
role in the causal chain leading to the act. Epidemiological investigations
are necessary to clarify the correlates of alcohol-related violence in
human populations and in various subgroups. Investigations of the alcohol-violence
relationship that are of particular interest are:
- Studies that disentangle the relationship between alcohol availability
and rates of suicide, homicide, assault, rape, and intimate partner
violence.
- Studies that investigate the risk and protective factors (e.g., personal, environmental,
genetic) for alcohol-related violence, and determine how these factors
vary and/or interact in different populations and/or are moderated by
factors such as family history, education, and socioeconomic status.
- Studies that clarify our understanding of how heavy alcohol use interacts with
impulsivity, negative affect, interpersonal disruption, and depression
to increase risk for suicide. Based on this theoretical foundation,
researchers should develop and test clinical interventions that address
co-occurring suicidality and alcohol dependence.
Studies of behavior, cognition and expectancies
Investigations of how alcohol consumption may alter behavior, including aggression,
by influencing the processing of social information are of particular
interest. These include:
- Studies of how alcohol-related behavioral expectancies regarding post
drinking aggression vary by age, sex, manner of drinking, drinking history,
pre-existing aggressive inclinations and personality traits (such as
impulsivity and dominance).
- Studies that investigate the mediating role of personality variables on the
relationship between alcohol consumption and violent behavior.
- Studies of the process of decision making in the context of alcohol-related
aggression under varying levels of provocation while controlling for
individual differences in alcohol expectancies and dispositional aggressiveness.
Availability
Changes in laws and other public policies--and the subsequent social changes
brought about by their adoption--provide "natural experiments"
through which the nature of alcohol-related violence
can be examined. Needed are:
- Investigations of the nature and strength of the relationship between
geographic density of public drinking places, off-premise alcohol outlets,
and violence.
- Studies that assess the impact of changes in laws regulating the availability
of alcohol on rates of alcohol-related violence.
- Studies of the ways in which bar clientele, bar management practices, and the
physical settings of bars affect the likelihood of bar violence.
D. CHRONIC DISEASES
MAGNITUDE OF THE PROBLEM
Chronic diseases include numerous diseases and conditions related to alcohol
consumption. The chronic condition most strongly related to alcohol is
liver disease. In 2001, 12,201 deaths from alcoholic liver disease and
6,719 deaths from unspecified liver cirrhosis were attributable to alcohol.
Together these 18,920 deaths accounted for more than half of all chronic
disease deaths attributable to alcohol as well as over half of the years
of life lost attributable to chronic alcohol conditions.
Aside from diseases of the liver, chronic diseases associated with alcohol consumption
include (but are not limited to) heart disease, cancer, and diabetes mellitus.
These three chronic diseases are respectively the first, second, and sixth
leading causes of death in the United States (Hoyert et al., 2005). Cardiovascular
disease accounts for approximately 45% of deaths in men older than 35
years and 37% of deaths in women older than 35 years (Gunzerath et al.,
2004). Cancer accounts for about 23% of US deaths, with the three most
common cancers, lung, breast, and colorectal cancers accounting for more
than one-third of cancer deaths (Jemal et al., 2005a). Almost 9% of Americans
age 20 and older have diabetes (National Diabetes Information Clearinghouse,
2004). Compared to non-diabetics, diabetics have double the risk of death.
DEPR is currently conducting under contract, a meta-analysis project to
estimate alcohol attributable fractions for chronic diseases including
heart disease, cancer and diabetes mellitus.
TRENDS
Liver Disease
During the past 20 years alcoholic liver disease deaths have declined 14% from
6.79 per 100,000 population to 5.86. Liver cirrhosis per 100,000 population
declined 27% during the same period from 9.99 to 7.25 (Figure 6).
Cardiovascular Disease
In 2002 the death rate for heart disease was 59 percent lower than the rate in
1950. The death rate for stroke declined 69 percent since 1950.
Cancer
A total of 1,372,910 new cancer cases and 570,280 deaths were expected in the
United States in 2005 (Jemal et al., 2005b). When deaths are aggregated
by age, cancer surpasses heart disease as the leading cause of death for
persons younger than 85. The death rate from all cancers combined has
decreased by 1.5% per year since 1993 among men and by 0.8% per year since
1992 among women. The mortality rate has also continued to decrease from
the three most common cancer sites in men (lung and bronchus, colon and
rectum, and prostate) and from breast and colorectal cancers in women.
Lung cancer mortality among women has leveled off after increasing for
many decades.
Diabetes
In 2003, more than 1.3 million adults between 18 and 79 years of age were diagnosed
with diabetes. From 1997 through 2003, the number of new cases of diagnosed
diabetes increased by 52% (CDC National Diabetes Surveillance System,
2005).
The incidence of diagnosed diabetes increased 43%, from 4.9 per 1000 population
in 1997 to 7.0 per 1000 population in 2003. Similarly, age-adjusted incidence
increased 41%, suggesting that the majority of the change was not due
to the aging of the population.
Figure 6
POPULATIONS AT RISK
Liver Disease
Men, older persons, and male White Hispanics are at higher risk for alcoholic
liver disease and unspecified liver disease. By gender, the age adjusted
rates for alcoholic liver disease in 2001 were 6.6 per 100,000 for men
and 2.2 for women (Yoon et al., 2004). For unspecified liver disease,
the comparable rates were 6.5 for men and 3.6 for women. Rates increase
after the age of 45.
Cardiovascular Disease
In 2001, the rate of death from heart disease was 31% higher among
blacks than whites and 49% higher among men than women. Major risk factors include high blood pressure high
blood cholesterol, tobacco use, diabetes, physical inactivity, and poor
nutrition. (
http://www.cdc.gov/nccdphp/burdenbook2004/toc.htm
).
Cancer
In analyses by race and ethnicity, African American men and women
have 40% and 18% higher death rates from all cancers combined than White men and women, respectively.
Furthermore, minority populations are more likely to be diagnosed with
advanced stage disease than are Whites (Jemal et al., 2006).
Diabetes
The estimated lifetime risk of developing diabetes for individuals
born in 2000 is 32.8% for males and 38.5% for females. Females have higher residual lifetime risks
at all ages. The highest estimated lifetime risk for diabetes is among
Hispanics (males, 45.4% and females, 52.5%). Individuals diagnosed as
having diabetes have large reductions in life expectancy (Narayan et al.,
2003).
RESEARCH CONTROVERSIES
Among the chronic diseases, liver diseases has the most
clear-cut association with alcohol. The association is less clear for cardiovascular disease and diabetes where low
levels of alcohol appear to be beneficial while higher levels are harmful,
and for breast cancer where moderate drinking may increase risk (Gunzerath
et al., 2004) . Evaluation of the dichotomy is complicated by numerous problems,
including the choice of a reference group (Fillmore et al., 2006), confounding
by lifestyle (diet, obesity, smoking, physical activity; Smothers and Bertolucci,
2001; Breslow et al., 2006), and measurement issues.
The US Dietary Guidelines (USDA, 2005) recommendation that
those who drink should do so in moderation is based, in part, on the apparent cardiovascular benefits
of moderate drinking. While this is not meant to encourage those who do
not drink to do so, there has been considerable media coverage on the
benefits of moderate drinking. Given that alcohol, even consumed in moderation,
could have negative consequences, including unintentional injury and harmful
interactions with prescription and over-the-counter medications, and for
women, elevation of breast cancer risk, it is important to determine whether
drinking in moderation is indeed beneficial and whether benefits outweigh
harms.
In epidemiologic studies, choice of the reference group
impacts the strength of observed risks. The majority of studies on moderate drinking and mortality have shown
a J-shaped curve with coronary heart disease (CHD) and total mortality
(driven by CHD mortality), with abstainers and heavier drinkers at increased
risk, and moderate drinkers at reduced risk (Fillmore et al., 2006). The
relationship persisted when former drinkers and persons with CHD were
removed from analyses. However, a recent meta-analysis (Fillmore et al.,
2006) found no protection for moderate drinkers when former drinkers (no
drinks in past year) were separated from complete abstainers and occasional
drinkers (drinks once a month or less) were separated from complete abstainers.
While this is a single study, albeit a meta-analysis, countering a large,
consistent literature finding the J-shape, it does suggest that the impact
of misclassification of drinkers on relative risks for CHD and total mortality
could be further considered.
Measurement is an important issue in alcohol research, both
in terms of alcohol and in terms of confounders. Most studies of alcohol consumption and chronic diseases
have measured alcohol as average volume, a measure that may not fully
explain important relationships between quantity of drinking, frequency
of drinking, and health outcome (Bondy, 1996; Rehm et al., 1996; Room,
2000). Average volume may obscure substantially different drinking patterns
due to quantity alone or frequency alone (Breslow and Smothers, 2004). Quantity
and frequency have been differentially associated with myocardial infarction
(Mukamal et al., 2003), high blood pressure (Russell et al., 1991), diabetes
(Conigrave et al., 2001), and body mass index (Bresland Smothers, 2005).
Studies of alcohol and chronic diseases that measure drinking patterns,
as well as the typically measured average volume, are needed. Drinking
patterns include quantity alone, frequency alone, stratified combinations of
quantity and frequency, and graduated frequency measures. For more on
these measurement issues, see the section entitled "Measurement of Alcohol
Consumption".
Confounders, including diet, obesity, smoking, and physical
activity, may obscure associations between alcohol and chronic diseases. In studies of associations
between alcohol and disease outcomes, measurement of confounders may be as
important as measurement of alcohol. For example, nutrition may be assessed
in numerous ways including dietary intake of foods and food groups, eating
patterns, nutrient intakes, nutrient biomarkers, and anthropometrics including
body mass index and waist circumference. Studies are needed to determine
the extent to which control for confounding can be improved by better
measurement.
RESEARCH PROGRESS
Liver Disease
It has long been established that alcohol consumption is a cause of
liver disease. Work which began in the 1920's has continued through several studies demonstrating
that heavy drinkers are at increased risk of death by liver cirrhosis
(see reviews by Schmidt and de Lint, 1972; Pell and D'Alonzo, 1973; Thun
et al., 1997; and Mann et al., 1993, 2004). Consistent with this, the
relative risk of cirrhosis morbidity rises with increasing levels of alcohol
consumption (Tuyns and Pequignot, 1984; Coates et al., 1986). In addition,
cirrhosis mortality rates are higher in societies where the average per
capita consumption of alcoholic beverages is higher (Lederman, 1956; Bruun
et al., 1975; Sk1980; Smart and Mann, 1991). This has led to the use of
cirrhosis mortality rates as a benchmark for evaluating the preventive
impact of policies to reduce the availability of alcoholic beverages (Bruun
et al., 1975; Edwards it al., 1994). In the US, alcoholic liver disease
mortality rates have declined since the 1970s. It has been suggested that
the declines may be due to increased participation in alcohol treatment,
decreases in alcohol consumption, and changes in consumption of certain types
of alcoholic beverage (Mann et al., 2003). However, the specific causes
are unknown and in need of further research. While it is clear that heavy,
long-term alcohol consumption can cause alcoholic liver disease in susceptible
people, not all heavy drinkers develop cirrhosis or alcoholic hepatitis.
This suggests the importance of additional risk factors including heredity, gender,
and diet.
Current research is investigating whether variation in
alcohol-metabolizing enzymes may affect susceptibility to alcoholic liver damage. ALDH and CYP2E1 variants are
of particular interest as polymorphisms in these genes have been associated
with liver disease. The fact that women develop cirrhosis at a lower cumulative
dose than men, combined with the fact that dietary intake and metabolic
processing of key nutrients, including fatty acids, differs by gender
suggests the need for gender-specific studies of genetic susceptibility,
taking nutritional status into account.
Recent research has also concentrated on alcohol's role in
mortality from hepatitis C (HCV). Heavy alcohol consumption accelerates the progression of Hepatitis C (HCV) (Poyard
et al., 1997, Schiff and Ozden, 2004). Heavy alcohol consumption also
interferes with the effectiveness of HCV treatment (Poyard et al., 1997).
Leading issues for future HCV research include the role of light and moderate drinking
in HCV progression, the relative contribution of alcohol consumption,
tobacco use, and nutrition in HCV progression, and the effectiveness of alcohol
cessation programs among HCV patients.
Cardiovascular Disease
Studies of cardiovascular disease and mortality consistently show a
J- or U-shaped curve among persons who are middle aged or older. As noted under ‘research controversies’
above, it is important to study and perhaps challenge the methodologic
conventions used in studies finding that relation, as well as consider
the measurement of alcohol and confounders. Measuring consumption separately
as quantity and frequency rather than the average volume, the measure commonly
used in epidemiologic studies appears to be important. Mukamal et al. (2003)
found that the protective effect of alcohol was more a function of frequency
of consumption than of volume; small amounts consumed several
times a week reduced risk of disease to a greater extent than the same
amount consumed over fewer occasions. A recent study by Mukamal et al.
(2005) suggested that alcohol consumption may be linked to potential hazards
among patients who survive acute MI. Although moderate intake has been
associated with lower mortality, binge drinking, even among light drinkers,
was associated with a 2-fold higher mortality. While the association between
alcohol and cardiovascular disease has been generally considered independent
of confounding by diet, a recent study (Breslow et al., 2006) suggests
that the way the confounding relationship is measured could affect study
outcomes. It is unlikely, due to ethical concerns, that a randomized,
controlled trial of drinking and CHD could be performed. Therefore, the best
possible observational epidemiologic data must be gathered to inform public
health policy. A recent NIAAA-sponsored workshop (Lucas et al., 2005)
suggested that epidemiologic areas needing research include determining
the potential benefits of alcohol in individuals who may be at increased
risk (genetic
and environmental) of developing cardiovascular disease, and confirming
the influence of beverage type and/or patterns of drinking on the relationship
between alcohol consumption and the risk of developing CHD.
Cancer
There have been numerous studies on the association between alcohol
and cancer (World Cancer Research Fund, 1997). Convincing evidence demonstrates that alcohol increases
the risk of cancers of the liver, mouth and pharynx, larynx, esophagus.
However, the evidence is less clear for lung, breast and colorectal cancers.
For all alcohol-related cancers, interactions between
alcohol and dietary components or nutrients impacted by alcohol (whether due to altered intake or metabolism) may
increase or decrease risk. Interactions between alcohol and nutrients that
are antioxidants or have antioxidant activity are important because ethanol
metabolism generates reactive oxidant species that damage tissue. These
nutrients include carotenoids, vitamins C and E, and the trace minerals
zinc, copper, selenium and manganese. Also of interest are methyl donors
including folate and choline, total energy intake, and dietary patterns.
Gene-environment interactions are particularly important given genetic
susceptibility for many cancers, polymorphisms in alcohol-metabolizing enzymes
and polymorphisms in nutrient-metabolizing enzymes. Gene-environment studies
of alcohol-related cancers are clearly needed.
Diabetes
Alcohol has a J or U-shaped relation with risk of type 2 diabetes
similar to that found for cardiovascular disease (Koppes et al., 2005). Moderate drinking appears to reduce
risk while heavier drinking appears to result in no increase or a modest
increase in risk. Obesity and physical inactivity are largely responsible
for the US epidemic of type 2 diabetes. About 65% of deaths among people
with diabetes are due to heart disease and stroke (CDC National Diabetes
Surveillance System, 2005). Analogous to cardiovascular disease, plausible
biologic mechanisms may explain the apparently beneficial effects of alcohol
on type 2 diabetes. However, confounding by lifestyle and other factors
could also be at least partly responsible for the effects seen. Diabetes
has a strong genetic component, making gene-environment studies important
in teasing out the role of alas an apparent protective factor. However,
methodologic studies should also be undertaken to determine whether observed
protective effects are potentially due misclassification of drinkers.
Because of the importance of obesity in the development of
diabetes and because of studies relating alcohol consumption to obesity, studies of the interactions
between alcohol and obesity leading to diabetes are warranted. Further,
studies of diets of children who drink – and whether interactions between
diet and alcohol contribute to their risk of obesity are important. Other
important areas include relations between alcohol and food cravings, satiety
and food intake, relationships between dietary patterns and drinking patterns,
and environmental variables influencing both food and alcohol intake.
As with cardiovascular disease, gene-environment studies would be beneficial
in determining the extent to which genetic susceptibility can
potentially be modulated.
OTHER AREAS OF IMPORTANCE
While this section has focused on diseases most strongly
associated with mortality outcomes, alcohol is associated with other
important diseases and conditions including, but not limited to: alcoholic lung disease, alcoholic cardiomyopathy, asthma,
osteoporosis, dementia, Alzheimer’s disease, depression, hypertension,
pancreatitis, and stroke. Further, alcohol impacts immune and endocrine
function, which in turn impacts risk for alcohol-related diseases including cardiovascular
disease and certain cancers. Epidemiologic studies of these other diseases,
including gene-environment studies, as well as the interactions between
biologic factors and alcohol in their development, are needed. Also important
are studies of the fetal origins of adult chronic diseases. Another area
of interest is how diet and nutrients interact with alcohol to influence
brain development, physiology and function.
RESEARCH QUESTIONS/ DIRECTIONS
- What are the causes of continuing high rates
of liver disease deaths among White Hispanic men?
- What is the contribution of alcohol use to
Hepatitis C?
- What is the role of light and moderate drinking in HCV
progression, the relative contributions of alcohol consumption,
tobacco use, and nutrition in HCV progression, and the effectiveness of alcohol cessation programs among HCV
patients?
- How do alcohol-nutrition interactions
contribute to the development of alcoholic liver disease?
Are drinking patterns associated with risk or
mortality for cardiovascular disease, cancer, or type 2 diabetes?
(i.e. quantity, frequency, satisfactions of quantity/frequency,
alcohol consumed at meals, alcohol not consumed at meals, heavy
episodic drinking).
Are drinking patterns associated with
dietary patterns that increase risk for cardiovascular disease,
cancer, or type 2 diabetes? More specifically do certain
types of drinking patterns lead to increased or decreased consumption of foods
or nutrients associated with greater risk?
Do certain types of drinking patterns
lead to changes in patterns of other risk factors for cardiovascular
disease, cancer, or type 2 diabetes?
Is there a relation between alcohol abuse and dependence and various chronic
diseases?
Is there a relation between early drinking onset and early binge drinking, and
chronic disease?
- Is there a relation of early onset of drinking and diet?
Can nutrition
and physical activity modify the development of alcohol-related chronic diseases?
- Which measures of nutrition and physical activity best minimize residual confounding
in studies of alcohol and chronic diseases?
What is the contribution of gene-environment interactions to the development
of alcohol-related
diseases, including cardiovascular disease, cancer, and diabetes?
Does alcohol consumption during pregnancy or breastfeeding contribute to the
origins of adult chronic disease?
How do diet and nutrients interact with alcohol to influence brain development,
physiology and function?
IV. KEY ISSUES OF ENVIRONMENTAL/ PSYCHOSOCIAL/CULTURAL CONTEXT
A. MINORITIES
MAGNITUDE
OF THE PROBLEM
Table 2 presents tabulations made for this report from the NLAES (1991-1992)
and NESARC (2001-2002) surveys. While we comment on alcohol use among
Asians/Native Hawaiians/Pacific Islanders and American Indian- Alaska
Natives, we must mention that NESARC did not oversample in these groups
as it did with Hispanic and African American populations. The prevalences
of heavier drinking, binge drinking, and alcohol dependence are substantially
elevated in the American Indian - Alaska Native population. Among Hispanic
men, the prevalences of alcohol dependence (5.9% in NESARC) and binge
drinking (23.2% in NESARC) were slightly higher than among White, Non-Hispanic
men. All three indices were higher in American Indian women and White-non-Hispanic
women than for women in the other groups reported.
TRENDS
Comparisons between the 2001-2002 NESARC estimates and the 1991-1992 NLAES estimates
in Table 2 present a mixed picture that does not lead to clear conclusions
regarding either trends in minority drinking habits or changes in intragroup
disparities over time. Among American Indian - Alaska Native men, heavier
drinking appears to have increased while dependence has fallen. Among
American Indian - Alaska Native women, heavier drinking seems to have
increased slightly, but binge drinking and dependence have both declined.
Among Hispanic men, binge drinking increased, but rates of dependence
fell. Among Hispanic women, rates of heavier drinking and dependence were
essentially stable while binge drinking rose somewhat. Prevalence trends
among White, non-Hispanics and for the total population also show a mixed
pattern of increases in some indicators but declines in others. Prevalence
trends increased across all indices for Asian/Native Hawaiian/Pacific
Islanders as a whole.
POPULATIONS AT RISK
High rates of heavy drinking, binge drinking, and alcohol dependence among minority
populations have long been of concern. However, in general population
survey data, such as that presented above, there are usually too few Native
Americans and Alaska Natives included in the sample to make precise estimates.
Similarly, there have been few studies of treatment or prevention protocols
developed explicitly for or adapted to Native American - Alaska Native
populations.
Table 2
Prevalence (%) of heavier drinking, binge drinking once a month or more,
alcohol abuse, and alcohol dependence in the past year, among total adult
population, by sex and race/ethnicity, United States, 1991–1992
and 2001–2002.
Sex and race/ethnicity1
|
Heavier drinking2
|
Binge drinking3
once a month or more
|
Alcohol dependence4
|
| |
NLAES |
NESARC
|
NLAES |
NESARC |
NLAES |
NESARC |
Both sexes
|
|
|
|
|
|
|
| White, non-Hispanic
|
9.90
|
11.29 |
11.73
|
13.00 |
4.35
|
3.83 |
| Black, non-Hispanic |
7.83
|
8.72
|
9.96
|
8.58
|
3.84
|
3.57 |
| American Indian/Alaska Native,
non-Hispanic |
10.59
|
12.75
|
17.09
|
14.43 |
9.01
|
6.35 |
| Asian/Native
Hawaiian/Pacific Islander, non-Hispanic
|
3.15
|
4.75 |
4.03
|
6.40 |
2.26
|
2.41 |
| Hispanic or Latino |
7.43
|
7.04
|
12.01
|
14.15
|
5.78
|
3.95 |
| Total |
9.28
|
10.26
|
11.36
|
12.38 |
4.38
|
3.81 |
| Male |
|
|
|
|
|
|
| White, non-Hispanic |
11.40
|
13.75 |
18.39
|
20.44 |
6.16
|
5.41 |
Black, non-Hispanic
|
11.16
|
12.45 |
16.38
|
14.95 |
5.86
|
5.09 |
| American Indian/Alaska Native,
non-Hispanic |
10.63
|
14.16 |
21.93
|
20.78 |
11.00
|
8.38 |
Asian/Native Hawaiian/Pacific
Islander, non- Hispanic
|
3.45
|
6.66 |
5.64
|
10.64 |
3.06
|
3.56 |
Hispanic or Latino
|
10.31
|
9.63
|
20.74
|
23.24 |
9.40
|
5.90 |
Total
|
11.01
|
12.81
|
17.99
|
19.81 |
6.33
|
5.42 |
Female
|
|
|
|
|
|
|
| White, non-Hispanic |
8.50
|
9.02
|
5.53
|
6.11 |
2.66
|
2.37 |
| Black, non-Hispanic |
5.14
|
5.82
|
4.79
|
3.62
|
2.21
|
2.39 |
| American Indian/Alaska
Native, non-Hispanic
|
10.55
|
11.46 |
13.12
|
8.66
|
7.38
|
4.49 |
Asian/Native
Hawaiian/Pacific Islander, non- Hispanic
|
2.83
|
2.98 |
2.30
|
2.45 |
1.41 |
1.34 |
Hispanic or Latino
|
4.52
|
4.36
|
3.21
|
4.72
|
2.15
|
1.94 |
| Total |
7.69
|
7.92
|
5.25
|
5.56 |
2.58
|
2.32 |
1 Total
includes other or unknown race/ethnicity.
2 Average daily
ethanol intake greater than one drink (0.6 oz) for women and greater than
two drinks (1.2 oz) for men.
3 Five or more
drinks in a single day.
4 With or
without abuse.
Source: NIAAA 1991–1992 National Longitudinal Alcohol Epidemiologic Survey
(NLAES) and 2001–2002 National Epidemiologic Survey on Alcohol and
Related Conditions (NESARC).
More data on Hispanics is becoming available. Results of recent surveys typically
show significant differences between men and women such that there is
concern about drinking and alcohol-related problems among men but less
concern about women. Whereas in past surveys, the magnitude of gender
differences in drinking patterns within the Hispanic population was relatively
large, these differences are diminishing. As noted in the "Liver
Disease" section above, there is a corresponding disparity in cirrhosis
mortality rates among Hispanic men and women.
Black drinkers who have alcohol use disorders are of concern with regard to health
disparities in access to treatment services and health insurance coverage
for such services. However, all the indicators in Table 2 show that heavier
drinking, binge drinking, and alcohol dependence remain lower in the Black,
non-Hispanic population than in the White, non-Hispanic population.
RESEARCH PROGRESS
Over the past several decades, epidemiologic research has documented substantial
variation in patterns of alcohol consumption and differential consequences
of that consumption across various racial and ethnic groups. Recent studies
have continued to indicate important differences within as well as between
ethnic and racial groups (e.g., Caetano, 2003; Spicer et al., 2003). We
have also learned that racial, ethnic, and cultural disparities in alcohol-related
problems vary with the problem under consideration. For example: the alcohol-related
death rate (i.e., for all categories of alcohol-related mortality combined)
is higher among African Americans than whites (NIAAA 1996); cirrhosis
death rates are very high among white Americans of Hispanic origin, lower
among non-Hispanic African Americans, and lower still among non-Hispanic
Whites (Stinson et al. 2001); alcohol-related traffic deaths are many
times more frequent (per 100,000 population) among American Indians or
Alaska Natives than among other minority populations (U.S. DHHS, 2000);
the incidence of fetal alcohol syndrome (FAS) is higher in some African
American and American Indian communities than in the general population
(Stratton et al., 1996); and recent increases in risky drinking behavior
(i.e., drinking and driving) have been reported among Hispanics (Voas
et al., 1998).
Problems related to alcohol are particularly elevated in Native American populations.
A study of Mission Indians found that 70% of men and 50% of women met
lifetime diagnostic criteria for alcohol dependence (Ehlers et al., 2004).
When compared to dependent individuals in the Collaborative Study on the
Genetics of Alcoholism Study (COGA), Mission Indian alcoholics were significantly
more likely to experience binge drinking, fights, and driving while intoxicated.
The incidence of fetal alcohol syndrome (FAS) in Native American populations
is also of great concern, which has led to the development of FAS diagnostic
tools that are applicable specifically to Native American groups (Hoyme
et al., 2005).
Intimate partner violence (IPV) among minorities has received recent research
attention. Significant ethnic differences exist in the prevalence of IPV,
and alcohol use is related to the occurrence of IPV in minority populations
just as it is in the general population. The role that alcohol use plays
in the occurrence of IPV appears to be stronger among Hispanics than among
Blacks or Non-Hispanic Whites (Field and Caetano, 2004).
RESEARCH QUESTIONS/DIRECTIONS
- Secondary analysis of existing data sets to broaden the array of available findings
on minority drinking patterns and problem prevalences. Many existing
data sets remain unexploited and could potentially yield important findings
on health disparities and intragroup differences in alcohol use and
alcohol use disorders.
- New studies oversampling Asians/Native Hawaiians/Pacific Islanders should
be conducted to better understand trends toward more alcohol use and
dependence. These groups are too small in ordinary general population
surveys for stable estimates of alcohol use and problem prevalence to
be made.
- NIAAA should partner with NIDA, NIMH, and NICHD to sponsor bi-national
(U.S./Canadian and U.S./Mexican) Native American - Alaska Native research
conferences. The goal should be to examine problems in these groups,
identify research priorities, and develop plans to initiate a program
of research and develop young investigators to carry out the work.
- Research aimed at developing improved prevention strategies for Native
American and Alaska Native populations, both broadly and as alcohol
relates to suicide, is needed.
- NIAAA should increase efforts to expand the pool of Asian/Native Hawaiians/Pacific
Islanders - Pacific Islander and American Indian - Alaska Native investigators.
B. AGING
MAGNITUDE OF THE PROBLEM
America's population of older adults is rapidly expanding. In the year 2000,
about 35 million Americans were age 65 years or older (U.S. Census Bureau,
2000). It is estimated that by the year 2020, 54 million, or one in six,
will be age 65 years or older.
Population-level data show that about 48% of older men (age 65 and older) and
32% of older women drink (Breslow et al., 2003). Most do so in moderation
as defined by the US Dietary Guidelines (USDA, 2005). Only 10%
of older men and 2.4% of older women are heavier drinkers (Breslow et
al., 2003). However, population-level data could obscure problem-drinking
occurring in certain populations. Studies suggest that a significant proportion
of "hidden" alcoholics may be age 60 and older (Cox et al.,
1997). The diagnosis of alcohol abuse and dependence can be difficult
in older people because its symptoms can be erroneously attributed to
other age-related medical or psychiatric conditions (e.g. depression,
insomnia, poor nutrition, and frequent falls) or to medication side-effects.
In addition, current DSM IV criteria for alcohol abuse and dependence
may not adequately assess alcohol problems in older adults due to the
absence of adverse social consequences that can be experienced by younger
at-risk drinkers (e.g., employment or legal problems) (Oslin, 2004; Fink
et al., 2002) and age-related physiological changes leading to tolerance
at lower consumption levels (Vestal et al., 1977). Hospital staff are
significantly less likely to recognize alcohol problems in an older than
in a younger patient (Curtis et al., 1989). In fact, physicians rarely
ask their older patients about alcohol consumption.
TRENDS
Results of cross-sectional (Breslow et al., 2003; Breslow and Smothers, 2004)
as well as longitudinal (Grant et al., 2004) studies have shown that older
adults consistently consume less than younger age groups. However, there
is a paucity of published, generalizable data on long-term trends in alcohol
consumption among older adults. For example, one study comparing survey
responses from two large national surveys showed increased prevalence
of alcohol abuse but not dependence among respondents over 65 years of
age from 1992-2002 (0.25 and 1.21 % respectively) (Grant et al., 2004).
Another longitudinal study (Moore et al., 2005) based on national survey
data found that while alcohol consumption declined with age from 1975
to 1992, consumption declined more slowly among more recent birth cohorts.
These results are suggestive of a potential shift in consumption patterns
among older adults as birth cohorts continue to age. Thus, these data
provide the basis for development of improved surveillance of consumption
patterns through the life course.
POPULATIONS AT RISK
Older persons differ biologically, psychologically and socially from younger
people, resulting in different health needs and health care utilization
patterns. Older men (and women of all ages) have a smaller volume of total
body water than younger men; therefore, they attain a higher blood alcohol
concentration from a given dose (Vestal et al., 1977). The greater vulnerability
of older persons to the effects of alcohol may be augmented by age-related
changes in functional status, nutritional status, and psychological and
cognitive status. Body composition continues to change with age; therefore
an 85 year old may be more vulnerable to alcohol’s effects than
a 65 year old (Vestal et al., 1977).
Alcohol also may affect the health of an older person by exacerbating sleep problems
(Brower and Hall, 2001), elevating blood pressure (Camargo et al., 1997),
and negatively affecting bone mineral metabolism (Ganry et al., 2001).
Alcohol use in older adults is also associated with hip fractures due
to falls (Rose and Maffulli, 1999) and other unintentional injuries including
automobile crashes (Higgins et al., 1996). The increased risk of hemorrhagic
stroke seen in the general population may be especially important in this
age group (Hillbom et al., 1996). Consumption of over one to two drinks
a day poses significant risks for cancer (Bagnardi et al., 2001), liver
cirrhosis, brain damage, and unintentional injuries (Rehm and Sempos,
1995).
The majority of older persons take medications, and alcohol interacts adversely
with many prescription and over-the-counter drugs. Studies indicate that
between 60 to 90 percent of elderly persons use some form of medication,
often more than one at a time (Chrischilles et al., 1992; Dufour and Fuller
1995; Pollow et al., 1994). Medications commonly taken by older people
which have a high potential for a negative reaction with alcohol include
analgesics, antihypertensives, anticoagulants, diuretics, antiarthritics
and psychoactive agents (Forster et al., 1993).
On the other hand, there is evidence of benefits from modest alcohol use. Moderate
alcohol consumption has been associated with increased survival in older
adults (Simons et al., 1996). Low to moderate consumption may offer some
protection against cardiovascular disease (CVD) , including ischemic stroke,
especially for those at moderate risk for CVD (Mukamal and Rimm, 2001)
and some studies have shown increased bone mineral density in postmenopausal
women who drink at moderate levels (Rapuri et al., 2000).
Not only is America's population graying, but cohorts of persons aging currently
(e.g. the so-called baby boomers) consumed more alcohol and other substances
during their early and middle years than the current cohort of elderly
persons. As a result, they may be more likely than current older adults
to continue to misuse alcohol as a coping mechanism in their later years
(Moos et al., 2004; Blow et al., 2004). In addition, we know that some
people increase their alcohol consumption later in life, often leading
to late-onset alcoholism (Atkinson et al., 1990). However, the etiologic
agents which produce late onset drinking have not been well described.
RESEARCH PROGRESS
Although studies have shown that the prevalence of excessive alcohol consumption
declines with age, a substantial proportion of older adults engage in
consumption patterns that exceed suggested guidelines (Moos et al., 2004).
In addition to the risk of adverse interactions with comorbidities and
medications, changes in life course conditions are believed to contribute
to increased risk for alcohol problems among older adults. These conditions
may include retirement-related unstructured free time and availability
of disposable income, gambling associated with binge drinking, or losses
due to death of loved ones (Oslin, 2004; Levens et al., 2005). Rates of
problem drinking vary widely depending on methods and definitions of alcohol
abuse and dependence (Moos et al., 2004; Blow et al., 2004; Oslin 2004).
Reported rates range from as low as 4% among representative longitudinal
samples to well over 50 % among smaller targeted community samples of
at-risk drinkers (e.g., in primary care and hospitals, geriatric mental
health clinics, nursing homes). A current study of older workers’
transition into retirement found that a shift from full-time work to bridge
employment and full retirement was associated with higher amounts of alcohol
consumption per drinking occasion (Bacharach et al., 2004). Other studies
of community-based samples in primary care have demonstrated the efficacy
of providers’ use of screening protocols to identify older adults
at risk for alcohol problems and reduction in at-risk drinking following
brief motivational interventions (Fink et al., 2002).
RESEARCH QUESTIONS/ DIRECTIONS
There is an important need for more epidemiologic information on alcohol consumption
by elderly persons; risks, benefits, predictors, and interactions with
chronic diseases of aging and their treatments. Future research should:
- Increase knowledge about the incidence, prevalence, etiology, course,
and natural history of alcohol-related problems, including, alcohol
abuse and alcohol dependence, among older adult populations, and among
subgroups of this population (e.g. racial/ethnic groups, men and women,
the very old, those in group housing).
- Examine consumption patterns (e.g., frequency, quantity, duration, beverage
type) to further clarify criteria for development of appropriate alcohol
consumption guidelines across the lifespan, including, relationships
between environmental, personal, socio-cultural and genetic factors
in the development and course of short- and long-term alcohol-related
problems among older adult populations.
- Examine the role of alcohol-nutrition interactions in alcohol-related disease
in the elderly.
- Test and refine screening protocols in primary care settings to assess alcohol
use problems and implement interventions to reduce alcohol consumption
to appropriate, safe levels.
- Elucidate the role of alcohol consumption in unintentional and intentional injury
(e.g., falls, vehicular crashes and suicide) and medical complications
from prescription and over the counter medication use (e.g. regular
vs. intermittent use, type of medication) in older adults.
- Examine the relationship of alcohol consumption and alcohol abuse and dependence
to the development, course and outcomes of other psychiatric disorders
and medical conditions (e.g., heart disease, cancer, liver disease and
degenerative brain disorders) among older adult populations.
C. HEALTH SERVICES
MAGNITUDE OF THE PROBLEM
In 2004 in the United States 15.2 million people were dependent on or abused
alcohol and 3.9 million were dependent on or abused both alcohol and drugs,
all total about 7.6% of the population 4 . These numbers have been relatively consistent since
2002 (SAMHSA, 2005).
An estimated 3.8 million people ages 12 and older 1.6% of the population received
treatment for a problem related to alcohol or drug use in 2004. Of those
1.2 million received treatment only for alcohol problems and 1.5 million
for both alcohol and drugs while 0.7 million for drug problems only. Changes
between 2003 and 2004 were not significant (SAMHSA, 2005).
Based on a survey date of March 31, 2002, SAMHSA determined that about 13,720
facilities provided services to about 1.13 million clients in treatment.
However, there were small absolute decreases in those numbers during
2003, down to about 13,623 treatment facilities providing care to about
1.09 million clients (SAMHSA N-SSATS). On the other hand, in 2001 SAMHSA
reported about 1.8 million substance abuse treatment admissions in the
U.S., for which about 436,142 (24.3 percent) were attributable to “alcohol
only” problems, and an additional 359,521 admissions (20.0 percent)
were attributable to “alcohol with secondary drug” substance
abuse problems. Those counts showed small absolute increases during
2002, when there were 1.9 million admissions, of which about 450,263 (23.6
percent) were “alcohol only” problems and about 367,740 additional
admissions (19.3 percent) were for “alcohol and secondary drug”
problems (SAMHSA TEDS).
TRENDS
Figure 7 shows that the percentage of substance abuse admissions for an alcohol
problem without a concomitant drug problem has declined during the past
decade. At the same time, the percentage of admissions for a drug problem
without a concomitant alcohol problem has increased and the percentage
of admissions for combined drug and alcohol problems has remained roughly
stable. It remains true that the great majority of admissions include
an alcohol problem, whether combined with a drug problem or not (Figure
8).
In the past decade substance abuse spending expressed as a percentage of total
health care spending decreased even more than other types of health care
spending. In 1992, substance abuse expenditures accounted for 1.16 percent
of total health care expenditures; this declined to 0.2 percent by 2001.
The percentage of privately-insured beneficiaries using any substance
abuse services declined by
23 percent from 1992 to 2001, from 0.64
percent of enrollees to 0.49 percent of enrollees. That decline was evident in all
categories: inpatient, outpatient, and
pharmaceutical utilization. Substance abuse spending per covered
life (in adjusted dollars)
dropped from about $21.16 in 1992 to
about $4.46 in 2002 (Mark and Coffey, 2004).
Figure 7
Substance Abuse Treatment Admissions, by Category of Primary Alcohol or
Drug Problem and by Year, Among Persons Who Were Admitted to Any Illicit
Drug or Alcohol Treatment in the Past Year, United States, 1993-2003
Source: SAMHSA, Office of Applied Studies, Treatment Episode Data Set (TEDS).
Figure 8
Substances for Which Last or Current Treatment was Received,
Among Persons Age 12 or Older Who Received Substance Use Treatment in the Past
Year, United States, 2004
Source: SAMHSA, National Survey on Drug Use and Health (NSDUH),
2004 Detailed Tables, Table 5.49A. Substance Dependence, Abuse and Treatment
At the same time, there has been a dramatic shift in substance abuse expenditures
away from private financing and toward public payers. There has also been
a shift away from hospital treatment settings and toward outpatient settings.
Overall, research evidence has indicated that prospective growth in substance
abuse expenditures had been contained and might remain constant relative
to the growth in other types of health care spending (Dilonardo et al.,
2000).
These trends have paralleled developments within the risk structure of health
insurance markets. Managed behavioral health care organizations, or so-called
“carve-out” arrangements have generally transferred economic
risk and accountability from the primary payor to the contracting vendor,
thereby diminishing overall quality of care, in the context of generally
reduced access to such services (Garnick et al., 2001). Overall, the largest
American firms are the most likely to engage in contractually “carving
out” behavioral health care services, suggesting on one hand that
employees of the largest firms might face diminished access (Hodgkin et
al., 2000).
POPULATIONS AT RISK
Age
Among older adults (ages 55 or older) admitted for substance abuse treatment
alcohol was the most frequently reported primary substance of abuse in
the years between 1995 and 2002. These admissions increased 19 percent
for men and 24 percent for women (1995: 33,100 men and 7,000 women; 2002:
39,300 men and 8,700 women) (SAMHSA TEDS, 2005). Among adolescents, between
1992 and 2002 substance abuse treatment admissions in which alcohol had
been reported as the primary substance of abuse decreased from 56 percent
to 20 percent of all adolescent substance abuse treatment admissions (SAMHSA
TEDS, 2004).
Mental health Comorbidities
NIAAA-supported research determined that the one-year prevalence rates of a diagnosis
of depression, alcoholism, and depression and alcoholism were 39.2, 2.0,
and 1.0 per thousand persons, respectively. Among persons with depression,
only 2.5 percent were identified as having alcoholism. Nevertheless, patients
with both depression and alcoholism utilized twice the number of mental
health and
substance
abuse visits over the two years after their diagnosis
and exhibited five times
the rate of psychiatric hospitalizations than patients who had depression
alone. Inpatient and outpatient mental
health and substance abuse expenditures were 80 percent and 68 percent higher among
patients with alcoholism and depression, respectively,
than among those with alcoholism alone. Overall, total medical care expenditures
for persons with both alcoholism
and depression were 44 percent higher ($7,323 greater) on average than
those for persons with depression alone over a
two-year period. Total inpatient costs were 59 percent higher ($3,341 greater) for persons
with both diagnoses, and total outpatient
costs were 46 percent higher ($3,878) (Mark, 2003).
At the same time, successful treatment for comorbid mental health conditions
assists successful recovery from alcohol dependence. Investigators in
a managed care organization examined the relationship between use of psychiatric
services and alcohol and drug treatment outcomes five years following
the completion of substance abuse treatment. They hypothesized that utilization
of psychiatric services would serve as a predictor of long term abstinence.
They determined patients who received a threshold level of psychiatric
services, at least 2.1 hours a year on average, were more likely to be
abstinent at five years than patients who received less than that threshold
level, such that utilization of psychiatric services among patients with
chemical dependency is associated with enhanced long-term outcomes (Ray
et al., 2005).
Chronic medical conditions
Substantial evidence exists that persons with diagnosable alcohol disorders exhibit
elevated rates of co-morbid conditions and yield higher health care costs,
but until recently little evidence had discerned whether those effects
could also be associated with “hazardous drinkers,” defined
as “those whose consumption increases their risk of physical and
psychological harm.” In a NIDA-funded study, investigators discerned
a hazardous drinking prevalence rate of 7.5 percent within a retrospective
sample of HMO primary care clinic patients. This prevalence was comparable
to that of hypertension and diabetes in the same HMO population. The investigators
suggested that screening and brief intervention at lower-threshold alcohol
abuse levels would detect both anticipated and unanticipated co-morbid
conditions sooner (Mertens et al., 2005).
RESEARCH PROGRESS
Recent research has focused on several means of improving care. These include
encouraging treatment practice improvements such as the use of screening
and brief interventions and the use of pharmacotherapy in treating alcohol
dependence. They also include moving toward a chronic care model by providing
continuing care and making use of self-help groups in the aftercare phase.
Investigators determined that the cost of a managed care organization-based screening
and brief intervention strategy is modest, and that it could be recommended
as a conventional prevention tool within such payment environments. They
compared a model in which brief interventions were delivered by a brief
intervention "specialist" to those where the brief intervention
was delivered by a "daily practitioner." Findings indicated
that the increased costs of both models were modest: $2.59 per patient
for the "specialist" model vs. $3.43 per patient for the "daily
practitioner" model (Mark, 2003).
Few physicians are prescribing Naltrexone as an adjunct to therapy. Investigators
in an NIAAA-supported study determined that the two main self-reported
reasons why physicians did not prescribe Naltrexone among their alcohol
dependent patients were (a) patients’ refusal to comply with prescribing
regimes (23 percent), and (b) patients’ inability to afford this
medication (21 percent). Physician perceptions of Naltrexone's effectiveness
and safety were significantly associated with their prescribing patterns.
Physicians who had more exposure to information about Naltrexone were
more likely to prescribe it (Mark et al., 2003).
In a study of the characteristics associated with successful continuing care
regimens, investigators found that a combination of brief, weekly telephone
contacts, and support group attendance during the first month following
outpatient completion, yielded outcomes that were as successful as those
generated by more intensive face-to-face continuing care interventions
(McKay et al., 2005).
In the report of key research findings from a 2001 consensus-development conference
on addiction-related self-help groups, a working group of national experts
recommended that all health care professionals “should have some
knowledge about how to refer patients effectively to [self-help] groups”
and that a credentialing program related to self-help group referral practices
might facilitate this outcome. They also recommended that government policies
work to encourage wider use of self-help organizations by underserved
populations, especially by making them more widely available to adolescents
and persons living in rural areas (Humphreys et al., 2004).
Other research has focused on the role of health insurance. These studies have
concentrated on changes in Medicaid policy, potential barriers to screening
and brief intervention, and the relationship between insurance status
and access to care.
In an NIAAA- and NIDA-funded study, researchers determined that the effects of
an overhaul of a state’s Medicaid system led to diminished outcomes
for substance abuse treatment. They found that a plan to create a modified
Oregon Health Plan that integrated substance abuse treatment service with
physical care for Oregon's Medicaid beneficiaries evolved in a multi-layered,
“carved-out” approach. Changes in the administration and management
of the chemical dependency benefit, financial losses by participating
health plans, and lack of training and incentives created poorer overall
treatment outcomes (Laws et al., 2002).
Another study reviewed the degree to which emergency physicians refrained from
delivering brief interventions due to concerns about patient confidentiality
and prospective denial of insurance coverage for an emergency department
episode. The investigators also surveyed state insurance commissioners
to determine whether state statutes allowed denial of coverage for injuries
sustained while impaired by alcohol. They determined that regulations
that would limit an emergency physician's freedom to deliver a brief intervention
existed in 38 states. They recommended segregating information about alcohol
use in a medical record and assigning substance abuse counselors to screen
all trauma patients, as the practical approaches toward ensuring the confidentiality
of alcohol information under current federal regulations (Rivara et al.,
2000). A follow-up survey identifying physician barriers, investigators
found that 27% of trauma surgeons perceived alcohol screening to be a
threat to reimbursement, 2% predicted that screening and intervention
would substantially increase health care costs, 7% considered screening
too time-consuming, and 14% expressed concerns that screening would compromise
patient confidentiality (Schermer et al., 2003).
A final study examined how different types of health coverage influence the likelihood
of entering alcohol treatment. The project studied the extent to which
persons in treatment were able to utilize their insurance resources to
help cover the costs of their care. The findings were that having private
insurance coverage did not significantly alter the odds of treatment entry
compared to being uninsured or being on Medicaid. Being in a private managed
care plan, as compared to traditional indemnity coverage, also did not
appear to affect the chances of treatment entry. Furthermore, 10% of those
who were privately insured reported having paid for all of their treatment
costs out of pocket. Thus, there are significant gaps between insurance
coverage and access to care (Schmidt and Weisner, 2005).
RESEARCH QUESTIONS/ DIRECTIONS
- Studies should examine the effectiveness and cost effectiveness of screening
and brief interventions delivered to adolescents and young adults. While
these strategies have shown great promise among adults, they may also
be of particular value in interrupting the developmental course that
leads from early onset of drinking toward the development of abuse and
dependence. Both primary care and emergency room settings should be
studied as sites for these interventions.
- Research should analyze the gap between the need for treatment in the general
population and the receipt of treatment. Research should determine whether
this gap is widening or narrowing over time and whether this gap is
especially pronounced among any age, gender, ethnic, or income groups.
- Studies should be conducted of the costs and cost-effectiveness of improved
services delivery. Given current financial constraints, improved treatment
techniques will need to demonstrate not only that they are more effective
but also that they are more cost-effective than standard treatment.
- Research should identify and quantify the sources of variation in access to
and utilization of alcohol-related prevention interventions and treatment
programs.
- Assessments should be made of the behavior of alcohol treatment providers. This
research should concentrate on identifying the kinds of provider training
and clinical supervision that yield quantifiably better treatment outcomes
and the training and supervision strategies that promote innovation.
D. PUBLIC POLICY
MAGNITUDE OF THE PROBLEM
A wide range of public policies may affect alcohol consumption and other behaviors
relating to alcohol, and therefore can have important influences on public
health outcomes. Many have not been subjected to extensive scholarly examination.
Public policy toward alcoholic beverages and alcohol-related behaviors
in the United States is complex, involving statutes, regulations, and
case law established at the Federal, State, and local levels of government.
Research efforts to identify the effects of specific policy interventions
on particular outcomes of interest confront challenges that are common
across much of the social science enterprise, including issues of measurement,
confounding, and multiple causal pathways, to name a few. The scientific
challenges are further complicated because of the complex ways in which
public policies influence particular aspects of alcohol-related behavior.
Thus, the overall effects of policy measures on health-related outcomes
may depend on the detailed characteristics of the causal pathways through
which policies work (Edwards et al., 1994; Babor et al., 2003).
In the United States, laws, regulations, and jurisprudence address (1) alcoholic
beverage production, packaging, transportation, marketing, taxation, sales,
and consumption, (2) financing and delivery of alcohol-related treatment
and preventive services, and (3) behaviors that may be affected by alcohol,
such as driving and boating. In addition, public policy may take in to
account the context in which behavior occurs, such as public vs. private
locations, in or near educational/correctional/transportation facilities,
the ages of the individuals involved, time of day, and other considerations.
Alcohol-related public policies and laws may be established by governments
at all levels (Federal, State, county, and municipal). However, following
the ratification of the 21st
Amendment to the United States Constitution, the primary authority to
regulate alcoholic beverages rests with the States (NIAAA APIS, 2006).
Scientific research has identified a number of alcohol-related policies that
have significant effects on public health outcomes. As noted in two of
our chapter sections on unintentional injury and intentional injury, among
the policies with the best evidence of effectiveness is the minimum legal
drinking age of 21, which has been shown to reduce consumption and alcohol-related
traffic crash deaths among youth 16-21. (Hingson and Winter, 2003; O’Malley
and Wagenaar, 1991; Shults et al., 2001; Wagenaar and Toomey, 2002). Although
every State now sets the minimum age for possession and purchase of alcoholic
beverages at 21, there is still a substantial degree of variation in State
policies toward underage drinking. Many States afford significant exceptions
to the laws against possession and consumption (e.g., except for in a
private residence, or only in a public place) (NIAAA APIS, 2006).
Alcoholic beverage taxes are another policy for which there is substantial evidence
of effectiveness. Although tax rates are most often established for fiscal
rather than public health purposes, a number of studies (but not all)
have found significant relationships between higher taxes on alcoholic
beverages and lower rates of traffic crash fatalities or drunk driving,
particularly among younger drivers or during nighttime hours. Other research
has found associations between higher alcoholic beverage taxes and lower
rates of some types of violent crime, reduced incidence of physical child
abuse committed by women, and lower rates of sexually transmitted diseases
and liver cirrhosis mortality, as well as with increases in college graduation
rates. Although considerable uncertainty persists about the magnitudes
of the relationships involved, the weight of evidence suggests that higher
taxes on alcoholic beverages are related to reductions in consumption
and in a range of adverse consequences associated with alcohol use (Chaloupka
et al., 1998; Cook and Moore, 2002; Kenkel and Manning, 1996; Leung and
Phelps, 1993).
Policies addressing drinking and driving have been shown to reduce traffic fatalities.
Every State has now established a law that makes driving with a blood
alcohol concentration of .08 percent a per se offense, meaning
that it is not necessary to present evidence of intoxication or impairment
to establish that a violation has occurred. All States have also adopted
so-called “zero tolerance” laws that set BAC limits for drivers
under the age of 21 at no more than .02 percent. These laws, combined
with a variety of other policies designed to deter driving after drinking,
have helped reduce alcohol-related traffic fatalities per population in
the United States by 50% since 1982 (Hingson et al., 1994; Hingson et
al., 1996; Hingson and Sleet, 2006; Shults et al., 2001).
Public policies affecting the delivery and financing of alcohol-related treatment
and preventive services may have important effects on access to treatment
services. Most States now require health insurers to cover treatment for
alcoholism, and a number of other States require such coverage to be offered
but not necessarily included in every insurance contract. However, no
Federal laws require coverage for alcoholism treatment, and Federal law
exempts many large employers from State laws making such requirements.
As a result, many insurance policies may not include coverage for alcoholism
treatment (Sing et al., 1998; NIAAA, 2006).
Many States have laws, known as “Uniform Accident and Sickness Policy Provision
Laws,” or UPPL, that permit health insurers to deny payment for
losses that are the result of the insured person being intoxicated. Researchers
have suggested that these policies create a disincentive for health care
providers in emergency and primary care settings to screen for alcohol
problems, with the result that fewer individuals in need of treatment
for alcohol problems are referred to treatment just when such referrals
might be most effective (Chezem, 2004; Rivara et al., 2000). While studies
have not yet established the true effects of UPPL provisions on treatment
referrals, a few States have recently enacted laws that prohibit exclusion
of insurance benefits on the basis of intoxication.
TRENDS
No broadly generalizable information on changes in alcohol-related public policies
has been reported. However, with respect to specific policy topics, the
following trends are clear:
- 1) Restrictions relating to drunk driving have become more stringent and penalties
more severe. All states have established a legal per se blood
alcohol limit of 0.08% and 41 states have adopted administrative license
revocation allowing laws immediate license seizure for persons above
the legal limit. All states have made it illegal for persons under 21
to drive after any drinking (Zero Tolerance Laws).
- 2) Restrictions relating to underage drinking and access to alcohol have become
somewhat more stringent over the past 25 years. This follows a period
of relaxation of these laws through the 1970s. All states have adopted
a legal drinking age of 21 making it illegal for those under 21 to purchase
or possess alcohol, and illegal for those 21 or older to sell or furnish
alcohol to persons under 21. However, a number of exceptions apply to
these general prohibitions, potentially weakening their effect. For
example, in 28 states, parents can legally provide alcohol to their
children, at least in some settings, and 11 states permit minors to
possess alcohol in some locations regardless of approval or supervision
by their parents.
- 3) Tax rates on alcoholic beverages have tended to decrease in real (inflation-adjusted)
terms over time, punctuated by sudden increases, most notably the 1990
increases in Federal excise rates on beer and wine and the 1985 and
1988 increases in Federal excise rates on distilled spirits.
- 4) Laws have increasingly encouraged or required health insurance coverage for
treatment of alcoholism. A number of States have repealed laws permitting
denial of health insurance benefits for injuries sustained as a result
of intoxication.
- 5) Long-term trends have favored increased legal availability of alcoholic beverages,
from the repeal of national prohibition, privatization of some State
monopoly operations, relaxation of restrictions on Sunday sales and
extended opening hours, and local zoning laws permitting alcoholic beverage
sales in broader categories of events and settings.
POPULATIONS AFFECTED
Public policies that plausibly affect alcohol-related outcomes address a broad
range of population groups. In many cases, the individuals or organizations
targeted by a particular policy are not the actors whose behavior is the
primary or ultimate target. For example, health insurance parity laws
place restrictions on health insurers in terms of the provisions that
may be included in health insurance policies, but the target behavior
is treatment-seeking and referral by individuals with drinking problems
and by physicians, employers, courts and law enforcement agencies, etc.
Public policies that affect alcohol consumption or related behaviors may affect
numerous specific outcomes. Many studies of the effects of specific policies
treat these behaviors (including consumption levels) as the outcomes of
interest. Dependent variables in such studies may include average consumption,
participation in (or frequency of) heavy (binge) drinking, participation
in (or frequency of) driving after drinking, etc.
Other studies examine the relationship between the presence or absence of a particular
policy and one or more downstream outcomes thought to be affected without
assessing changes in consumption or other intermediate alcohol-related
behaviors. Traffic crash fatality rates are probably the most commonly
studied of such downstream outcomes. A growing literature has studied
the effects of changes in alcoholic beverage prices or taxes on such other
outcomes as cirrhosis rates, educational outcomes, violent crime, domestic
violence, and sexually transmitted diseases.
In principle, policies that affect consumption may influence any and all alcohol-related
outcomes. However, it is likely that certain policies have stronger effects
on some categories of consumption than others (for example, frequency
of heavy consumption, or maximum number of drinks per occasion, or drinking
by potential drivers, or drinking by relatively light drinkers) and thereby
have stronger associations with some outcomes rather than with others.
The issue of such differential effects has received relatively little
research attention to date, and represents a key area for future research.
RESEARCH PROGRESS
A few categories of alcohol-related policies have received significant scientific
study, while most others have received little if any research attention.
In addition, most studies have examined one or two specific policies at
a time, without detailed consideration of possible interactions among
policies. Among the policy areas that have received the greatest research
attention are minimum drinking age laws, taxation of alcoholic beverages,
drinking driving laws (especially blood-alcohol concentration limits),
restrictions on alcohol availability, and, to a lesser degree, laws facilitating
access to treatment and restrictions on advertising and marketing.
The recent volume Alcohol: No Ordinary Commodity (Babor et al. 2003) reviews
evidence for a range of public policy interventions from an international
perspective. The stated purpose of the book is “… to describe
recent advances in alcohol research that have direct relevance to alcohol
policy on the local, national, and international levels.” Similar
reviews of policy effects include Cook and Moore (2002) and Edwards et
al. (1994). Hingson et al. (2006) have recently reviewed alcohol policies
that have reduced injury.
The weight of evidence from numerous studies suggests that beverage taxes may
be among the most effective of policy measures at reducing consumption
and ameliorating a variety of adverse outcomes. Although there is broad
consensus that higher prices lead to reductions in consumption, there
is significant variation in the magnitude of estimated effects, and some
of this variation may be attributable to methodological challenges that
confront research in this area (Leung and Phelps, 1993; Kenkel and Manning,
1996; Chaloupka et al., 1998; Cook and Moore, 2002). Some recent research
has begun to examine the empirical relationship between alcoholic beverage
taxes and prices (Young and Bielinska-Kwapisz, 2002; Kenkel, 2005). A
few analyses have explored how prices may exert differential effects on
consumption by moderate and heavy drinkers (Manning et al., 1995; Kenkel,
1996). A number of studies have reported significant effects of beverage
taxes on traffic crash fatalities (Saffer and Grossman, 1987; Chaloupka
et al. 1993; Kenkel 1993; Ruhm 1996). However, several recent studies
have found effects of taxes on traffic crash fatalities to be insignificant,
smaller than previously estimated, or unreliable (Dee, 1999; Mast et al.,
1999; Young and Likens, 2000). Methodological factors may help to explain
these differences in findings. Chief among these factors is the possibility
of unmeasured State-specific factors that influence both tax rates and
alcohol-related outcomes. Clearer understanding of the mechanisms through
which beverage taxes affect alcohol consumption and related behaviors
remains an important area for further research.
Other research supports the conclusion that minimum drinking age laws are effective
in reducing alcohol consumption and alcohol-related problems among young
people (Shults et al., 2001; Wagenaar and Toomey, 2002). In addition,
Chisholm et al. (2004) estimated the relative cost-effectiveness of several
interventions (taxation of alcoholic beverages, drunk driving legislation
and random breath testing, reduced hours of sale, advertising bans, and
brief physician interventions), concluding that both individual-based
interventions (e.g., brief physician advice) and population-wide interventions
(e.g., taxation) can improve public health, but that the relative cost
effectiveness of alternative approaches varies across countries with patterns
of consumption and problems associated with alcohol.
Because of the complexity of alcohol-related public policies, researchers face
the challenge of identifying how specific policy measures affect health
outcomes. The Alcohol Policy Information System (APIS), developed by the
National Institute on Alcohol Abuse and Alcoholism (NIAAA), provides authoritative,
detailed, and comparable information on alcohol-related public policies
in the United States at both the State and Federal levels. The APIS Web
site ( http://alcoholpolicy.niaaa.nih.gov
) provides public access to detailed information
on a wide variety of alcohol-related public policies. Intended primarily
as a tool for researchers, APIS features compilations and analyses of
alcohol-related statutes and regulations designed to simplify the process
of ascertaining the state of the law for studies on the effects and effectiveness
of alcohol-related policies.
APIS currently provides information on 36 policy topics in several formats and
at varying levels of detail. This information includes summary descriptions,
maps, detailed comparison tables, specific dates on which provisions became
or ceased to be effective, and excerpts providing the full text of relevant
passages of statutes and regulations. For most policy topics, APIS coverage
begins January 1, 1998. A list of current APIS policy topics is provided
in Table 3; additional details are available through the APIS web site.
RESEARCH QUESTIONS/ DIRECTIONS
- How do various policies relating to underage
drinking affect adverse alcohol-related
outcomes?
- Are there categories of policies in which
enforcement and other aspects of implementation
are especially critical in contributing to
desired outcomes?
- To what extent can repeal of “UPPL” laws
contribute to increased access to treatment, and
how much would that improve alcohol-related
health outcomes?
- How do restrictions on physical availability
(e.g., restrictions on days and hours of sale;
limits on numbers of outlets; and limitations on
brand selection and package sizes) and beverage
taxation compare in effectiveness at reducing
alcohol-related problems?
- Through what mechanisms, and among which
population groups, do taxes and other policies
that are effective in reducing adverse
alcohol-related outcomes operate? Do they reduce
the prevalence of consumption, the frequency of
consumption, the maximum number of drinks per
occasion, the frequency of heavy drinking
occasions, or other aspects of drinking related
behavior, such as the propensity to drive after
drinking or to engage in risky sexual behaviors?
Are heavier drinkers more or less susceptible to
policy effects than more moderate drinkers? Does
this relationship vary across ages,
socioeconomic classes, or other discernable
categories?
Table 3
Current APIS
Policy Topics
Alcohol Control Systems: Retail
Distribution Systems for Wine
Alcohol Control Systems: Wholesale
Distribution Systems for Wine
Alcohol Control Systems: Retail
Distribution Systems for Beer
Alcohol Control Systems: Wholesale
Distribution Systems for Beer
Alcohol Control Systems: Retail
Distribution Systems for Spirits
Alcohol Control Systems: Wholesale
Distribution Systems for Spirits
Keg Registration
Beverage Service Training and
Related Practices
Bans on Off-Premises Sunday Sales
|
4. Underage Drinking and Access to
Alcohol
Underage
Possession of Alcohol
Underage Consumption of Alcohol
Underage Purchase of Alcohol
Furnishing Alcohol to Minors
Minimum Ages for On-Premises Servers
and Bartenders
Minimum Ages for Off-Premises
Sellers
Loss of Driving Privileges for
Alcohol Violations by Minors
("Use/Lose” Laws)
Hosting Underage Drinking Parties:
Criminal Liability
False Identification for Obtaining
Alcohol
|
Alcohol
Beverage Taxes: Beer
Alcohol Beverage Taxes: Spirits
Alcohol Beverage Taxes: Wine
Alcohol Beverage Taxes: Sparkling
Wine
Alcohol Beverage Taxes: Coolers
|
5. Health Care
Services and Financing
Insurers’ Liability for Losses Due
to Intoxication ("UPPL")
Health Insurance Parity for
Alcohol-Related Treatment
|
3. Transportation,
Crime, and Public Safety
Open
Containers of Alcohol in Motor
Vehicles
Blood Alcohol Concentration Limits:
Adult Operators of Noncommercial
Motor Vehicles
Blood Alcohol Concentration Limits:
Youth (Underage Operators of
Noncommercial Motor Vehicles
Blood Alcohol Concentration Limits:
Recreational Watercraft
Vehicular Insurance Exclusion
|
Alcohol &
Pregnancy: Mandatory Warning Signs
Alcohol & Pregnancy: Limitations on
Criminal Prosecution
Alcohol & Pregnancy: Civil
Commitment
Alcohol & Pregnancy: Priority
Treatment
Alcohol & Pregnancy: Legal
Significance for Child Abuse / Child
Neglect
Alcohol & Pregnancy: Reporting
Requirements
|
E. ADVERTISING
MAGNITUDE OF THE
PROBLEM
Alcohol messages and the modeling of
drinking behavior are ubiquitous in contemporary
American society. One part of this is the
advertising of alcoholic beverages. Alcohol ads
appear in print media, on television, on the radio
and on billboards. In addition, alcohol is promoted
through product placements, sponsorships, Internet
advertising, point-of-sale materials, and items with
brand logos.
Media exposure to
alcohol other than in advertising is also
widespread. Apart from direct advertisements,
drinking is also widely portrayed in television
programming and is regularly shown in movies for
persons of all ages. Nielsen research from 1998
estimated that American children watch on average 21
hours of television a week (Villani, 2001). Many
programs have alcohol content, which is most often
neutral or positive, and only rarely negative. The
same is true for movies the vast majority of which
depict drinking. A recent study (Sargent et al.,
2006) found an association between exposure to movie
alcohol use and early onset of drinking among
teenagers.
TRENDS
In 2001, alcohol companies spent $4
billion to advertise and promote their products. Of
this $1.57 billion was in the traditional measured
media which includes television, radio, print and
outdoor advertisements and the rest was for
unmeasured promotions such as sponsorships, Internet
advertising, point-of-sale materials, product
placement, items with brand logos, and other means (CSPI,
2003). Growth in measured alcohol advertising since
1975 has been 20% greater than can be accounted for
by inflation.
The Center on
Alcohol Marketing and Youth (CAMY) monitors the
scope and extent of underage youth to alcohol
advertising in magazines, on television and through
other media. In the years from 1998 to 2002,
expenditures for beer advertising on television
increased 45% (to 972 million dollars) while
expenditures on liquor advertising increased 530%
(to 18 million dollars; CSPI, 2003).
POPULATIONS AT RISK
Who sees/hears alcohol advertising
and who is affected by it is an important issue.
While almost all persons are exposed to significant
amounts of alcohol advertising, certain populations
may be at particular risk in terms of “overexposure”
and/or vulnerability to alcohol’s effects, for
example youth, minorities and dependent drinkers.
Further some groups such as underage drinkers and
members of certain ethnic populations may be
“targeted” by certain advertisers. Others such as
dependent drinkers, or those in recovery, for whom
alcohol ads may provide drinking cues or triggers,
may be especially vulnerable to advertising. A
recent study comparing teens with and without
alcohol use disorders (AUD) found that teens with
AUD showed substantially more brain activation to
pictures of alcoholic beverages than controls (Tapert
et al., 2003).
Youth
The amount and
types of alcohol advertising youth experience and
its impact on their decisions about drinking are
important concerns. Alcohol advertising on
television is pervasive, often appealing to youth,
and shown at times when many youth are likely to see
it. For example, half of televised beer ads air on
Saturday or Sunday afternoons during sporting events
when they are very likely to be seen by youth
(Snyder et al., 2000). Youth are also routinely
exposed to print ads for alcoholic beverages, in
magazines, on billboards and on the Internet.
Descriptive research from the Center on Alcohol
Marketing and Youth (CAMY, 2002) found that youth
saw 45% more beer, 27% more distilled spirits ads,
and 60% more “malternative” magazine ads than adults
and that youth 12-17 were also “overexposed” to
alcohol ads on the radio relative to adults 21 years
and over, in the majority of markets monitored (CAMY,
2004). Further, radio alcohol ads were frequently
placed on stations with youth formats and were aired
when youth were most likely to be listening (CAMY,
2003). A study of internet use by youth found the
alcohol websites contained features appealing to
youth, such as video games and cartoons, but few
effective mechanisms to keep underage youth from
accessing the websites (CAMY, 2004).
Minorities
The literature on minority population exposure
to advertising is not as well developed as it is for
youth exposure to advertising. However, there is an
emerging interest in understanding the targeting of
products like tobacco and alcohol to minorities
(Gardiner, 2004) and examining the exposure that
comes from a greater geographic density of outlets
and large outdoor ads, which have been found
disproportionately in African American and Hispanic
neighborhoods (Hackbarath et al., 2001).
RESEARCH PROGRESS
The majority of
studies of the effect of alcohol advertising on
attitudes about drinking and on actual consumption
have been done in youth populations. Some research
with youth has directly studied young people’s
reactions to alcohol ads, and correlates of those
reactions. For example, a study of third, sixth, and
ninth graders showed that the third-grade children
who found alcohol ads desirable were also more
likely to see positive benefits from drinking and
desire products with alcohol logos. Older children
(ninth grade) in the study who found the ads and
logo products appealing were more likely to already
be engaged in drinking behaviors (Austin and Knaus,
2000). On the other hand, Zogg et al. (2004;
supported by AA 12128) in a study of perceived
positive and negative outcomes of alcohol use among
8th and 10th grade students
found no predictive effects of exposure to televised
alcohol advertisements, televised sports (which is
dense in alcohol advertising) or first-hand
observation of others drinking. More recently,
Austin et al. (2006), using a complex information
processing model found that media portrayals
influence children’s drinking through a progressive
decision-making process. They also found that the
interpretation of messages is at least as important
as media exposure per se, and that decisions are
both logic and affect based and incorporate inputs
from multiple sources including
media and parents.
Another group of
studies has examined the relationship of advertising
exposure and drinking outcomes. While two recent
cross-sectional studies found positive associations
between advertising and consumption (Collins et al.,
2003; Chen et al., 2005), results of earlier
cross-sectional studies that examined the
relationship between liking alcohol advertising and
current and future intentions to drink were mixed
(Kelly and Edwards, 1998; Wylie et al., 1998). The
results of a longitudinal study of New Zealand youth
indicated that liking of alcohol advertising at age
18 was related to higher levels of beer consumption
at age 21 (Casswell and Zhang, 1998). While there
are some methodological problems, as there are in
many of these studies, two prospective studies
indicate a positive relationship between exposure to
advertising and consumption. Ellickson et al. (2005)
found in a sample of 7th grade drinkers
and non-drinkers from North Dakota that several
forms of advertising predicted future adolescent
drinking for both groups (AA12127) and Stacy et al.
(2004) found that exposure to advertising increased
the risk of subsequent beer consumption.
In another recent
study, Snyder et al. (2006) interviewed a sample of
youth aged 15 to 26, from 24 Nielsen media markets,
on four occasions over a period of 21 months about
their drinking. Advertising exposure in the study
was measured both subjectively in terms of reported
exposure and objectively in terms of advertising
expenditures. Results indicated that each additional
advertisement seen increased the number of drinks
consumed in the past month by 1%. Further, youth in
markets with greater advertising expenditures drank
more: for each additional dollar spent per capita,
the number of drinks consumed per month increased by
3%.
Another group of potentially informative investigations are econometric studies
of the relationship between alcohol advertising and consumption. Results
of these studies have also been mixed. A study by Saffer and Dhaval (2002)
found that advertising increased consumption while others found that alcohol
advertising affects brand choice but not overall consumption (Nelson and
Morgan, 1995; Gius, 1996). Another recent study by Saffer (2003) suggests
that a complete ban on alcohol advertising might reduce the prevalence
of monthly drinking by adolescents, 12-18 years of age, from about 25%
to 21% and of binge drinking from 12 % to 7%. However, while comprehensive
advertising bans are potentially effective in reducing youth consumption,
they are not likely to receive public support and partial bans are likely
to result in the industry increasing their ads in other media (Saffer,
2002). A more recent study (Hollingworth et al., 2006) reached a similar
conclusion noting that advertising bans have the potential to significantly
reduce alcohol-related mortality and years of lost life due to premature
alcohol-related deaths, but also that a complete media ban is unlikely
since the risks of low levels of alcohol consumption are low.
In sum, results of research on the impact of alcohol advertising on actual drinking
behavior by youth has been mixed and observed effects have been small.
Furthermore many studies have been cross-sectional making it difficult
to draw definitive conclusions about the relationship between advertising
and alcohol consumption.
RESEARCH QUESTIONS/ DIRECTIONS
- We need more definitive studies of the fundamental question in this area which
is whether exposure to alcohol advertising influences alcohol consumption.
We need to understand the parameters of advertising that affect this
relationship such as type of exposure, context of exposure and type
of content as well as the aspects of drinking such as frequency of consumption,
volume of consumption and context of consumption that are affected.
- Some studies have focused on expectancy and intention to drink as intermediate
outcomes. While we need enhanced understanding of the processes by which
advertising affects these outcomes, we also need greater understanding
of the processes by which advertising affects the most important and
more distal outcome of actual consumption.
- We need a better understanding of how advertising affects decision making about
drinking among young people during a time of life when people are inclined
to engage in more risk behaviors, including alcohol consumption, and
a developmental period during which they may be especially vulnerable
to alcohol’s effects and adverse consequences.
V. OVERARCHING METHODOLOGICAL ISSUES
This final section covers overarching
methodological issues that affect more than one of the substantive areas
described above. Because these issues are crosscutting, we use a different
format than that used in the preceding sections, presenting first a description
of the methodological issues and then a set of research recommendations.
A. MEASUREMENT OF ALCOHOL CONSUMPTION
METHODOLOGICAL ISSUES
An important goal of alcohol epidemiology is to link alcohol consumption with
alcohol-related problems. At the level of a large population, analyses,
such as those that compare variation in per capita alcohol consumption
and mortality rates over time, are useful in demonstrating links between
consumption and its sequelae. Consumption data for such aggregate analyses
typically are based on information about alcohol sales or shipments. In
the United States, these data represent the standard against which other
estimates of alcohol consumption are based.
In contrast, analyses that link drinking behavior with related outcomes at the
individual level generally rely on survey data. Surveys of consumption
allow researchers to ask individuals about their drinking patterns and
to obtain other potentially related information, such as sociodemographic
characteristics, health status, and alcohol-related experiences. This
approach enables investigators to link alcohol consumption with various
outcomes at the aggregate level as well as to adjust for other aggregate
characteristics that might confound the associations being studied. In
addition, survey data permit researchers to identify abstainers and to
separately examine the impact of drinking frequency and drinking quantity.
Finally, survey data allow for tracking of specific patterns of risk drinking,
generally defined as drinking at a level that might result in psychomotor
impairment.
Because of the importance of survey data for estimating relationships between
drinking and alcohol outcomes, the general approaches and specific questions
used to assess alcohol consumption have received much attention (Alanko,
1984; Armor and Polkich, 1982; Midanik and Harford, 1994; Rehm, 1998;
Room, 1990). The National Institute on Alcohol Abuse and Alcoholism has
had a strong and continuing involvement in the measurement of alcohol
consumption as evidenced by the convening of conferences (Dawson and Room,
2000; NIAAA, 2003) and publications (NIAAA, 2004; 2005) directed to both
researchers and the clinical community. These developments are summarized
below.
THE DATA COLLECTION REFERENCE PERIOD
The first issue faced in designing an alcohol consumption survey is to decide
the reference period for which consumption data are collected (e.g., the
past year, past month, past week, or most recent drinking occasion). With
short reference periods, (i.e., one week or less) respondents can be asked
to describe the exact number, size and type of drinks they consumed each
day. This approach, referred to as exact recall, is thought to minimize
problems with memory loss and avoids the problems inherent in trying to
describe a respondent’s “usual” pattern of intake. Despite
these advantages, exact recall procedures are associated with significant
limitations. First, the short recall period may not accurately represent
the respondent’s typical consumption throughout the year, particularly
in populations where drinking volumes or patterns vary according to season
or are influenced by holidays. Second, the exact recall approach is not
well suited to populations where many drinkers consume alcohol on an infrequent
or irregular basis. In these cases, an exact recall approach is likely
to misclassify many infrequent drinkers as abstainers even though it may
accurately estimate the volume of consumption at the population level
(assuming that a representative week is selected). Third, a short recall
period generally is inadequate for simultaneously assessing alcohol-related
problems, many of which occur rarely and can be measured with sufficient
precision only over a longer period of time, typically a year.
A longer reference period (typically one year) is recommended for assessing both
drinking behavior and problems in countries such as the United States,
where many people are light, irregular drinkers and where large-scale
surveys assessing both alcohol consumption and related problems permit
individual level linkage of both types of data. With such a reference
period, respondents cannot be asked to recall each drink they consumed
during that time. Instead, researchers have developed various approaches
to estimate the respondents’ usual consumption. The challenge with
these approaches is how best to collect information that can simultaneously
yield accurate estimates of drinking frequency, volume, quantity, and
variability, as well as the prevalence of risk drinking. Over the course
of several decades of nationwide alcohol surveys conducted in the United
States, two general ways of obtaining summary consumption data have evolved
–the quantity/frequency (QF) approach and the graduated frequency
(GF) approach (Room, 1990; Greenfield, 2000).
The Quantity/Frequency and Graduated Frequency Approaches
In most surveys, the QF and GF questions pertain to consumption in the past year,
which is typically thought to reflect the respondent’s current drinking
status. However, these approaches can also be applied to other reference
periods, such as the past month, which is the time-frame most often recommended
for use with adolescent drinkers, or they can be designed to cover various
life stages or periods of heavy drinking e.g. the period of heaviest drinking
or to various life stages. In general, investigators have devoted less
attention to measuring lifetime consumption (Lemmons et al., 1997; Russell
et al., 1997; Skinner and Sheu, 1982; Sobell and Sobell, 1992) than to
measuring current consumption.
In its most basic form, the QF approach measures alcohol consumption with two
questions that inquire about (1) the overall frequency of alcohol consumption
within the reference period, and (2) the usual number of drinks consumed
on days when the respondent drank alcohol. The variable “usual number
of drinks” theoretically measures the most commonly consumed quantity
(i.e., the mode). Although past research suggests that responses actually
may reflect a quantity somewhere between the mode and the mean (Gruenewald
et al., 1996).
The GF approach, which since 1979 has been used in varying forms in the National
Alcohol Surveys, conducted by the Alcohol Research Group, asks respondents
how often during the designated reference period they drank various quantities
of standard drinks. In contrast to the QF approach, the GF approach provides
a standard set of drinking pattern measures (i.e., the quantities of drinks
for which frequencies are reported are the same for all respondents),
thereby facilitating the analysis of drinking patterns, estimation of
risk curves, and presentation of results.
The Beverage-Specific Approach
The basic QF and GF structures lend themselves equally well to questions on overall
alcohol consumption or the consumption of individual types of beverages.
Past studies have consistently shown that data from beverage-specific
questions, when summed across beverages, yield higher estimates of consumption
than data from a single series of questions on overall consumption of
all types of alcoholic beverages (Dawson, 1998; Russell et al., 1991).
Investigators cannot simply add drinking frequencies across beverages
to estimate overall drinking frequency because respondents may consume
more than one type of beverage per day. In order to collect optimal data
on both volume and pattern of drinking, surveys should include both beverage-specific
questions and questions on overall consumption. Questions about overall
consumption need not be asked in comparable detail to the beverage-specific
questions but should at least contain questions on overall frequency of
consuming any alcohol and overall frequency of consuming 5+ drinks, or
a similar indicator of risk drinking (e.g. frequency of being intoxicated
or feeling the effects of alcohol).
Precoded Response Categories
Alcohol surveys typically provide the respondents with precoded response categories
representing frequency of drinking rather than asking for the actual number
of drinking days. For example, research has shown that respondents find
it easier and less embarrassing to report a frequency of “once a
week” than “52 times”, at least when reporting sensitive
information, such as frequency of heavy drinking in the past year (Ivis
et al., 1997). The order of response categories can influence the accuracy
of the respondent’s answers. Ordering response categories so that
the highest frequencies are at the top of the list helps to make higher
frequencies seem more normal and less embarrassing to the respondent (Dawson
and Room, 2000). A disadvantage of precoded response categories is that
they limit the number of possible responses, especially in the upper ranges
of frequencies. Questions on the usual and largest quantity of alcohol
consumed often are asked in an open ended format.
Definitions of Drinking Status
Most surveys aimed at determining past year consumption begin with a short series
of questions intended to ascertain the respondent’s drinking status.
Based on these introductory questions, respondents generally are classified
into one of three categories of drinking status – lifetime abstainer,
former drinker, and current (i.e. past year) drinker. Distinguishing lifetime
abstainers from former drinkers is particularly important in epidemiological
research because any health benefits of light to moderate drinking (e.g.,
reductions in coronary heart disease) (Rimm et al., 1996) may be exaggerated
when former drinkers, who may have been heavy drinkers who stopped drinking
because of adverse health effects, are included in the category of abstainers
(Shaper, 1995).
Overall Alcohol Consumption
To determine a respondent’s overall alcohol consumption, researchers must
create analytic measures that describe drinking pattern and volume. At
a conference devoted to measuring alcohol consumption, participants recommended
the following items for presentation and analysis of consumption data
(Dawson and Room, 2000):
~Drinking status
~Volume of pure alcohol (i.e. ethanol) consumed
~An indicator of the frequency of risk drinking (i.e., drinking at a level that
might result in psychomotor impairment, such as frequency of drinking
5+ drinks)
5
.
Binge and Risk Drinking
Drinking too much, too fast and/or too often results in risks to drinkers and
those around them. Various definitions of binge drinking have been used;
most incorporated drinking 5 or more drinks on an occasion or in a row
for men and 4 or more for women. In 2004, the NIAAA National Advisory
Council approved the following definition which also incorporates a two
hour time frame: A “binge” is a pattern of drinking that brings
blood alcohol concentration (BAC) to 0.08 gram percent or above. For the
typical adult, this pattern corresponds to consuming 5 or more drinks
(male), or 4 or more drinks (female) in about 2 hours. Binge drinking
is clearly dangerous for the drinker and for society (NIAAA, 2004).
NIAAA also recently published a clinicians guide for primary care and mental
health clinicians which described risky drinking for men as 5 or more
standard drinks a day (or 15 or more per week) and 4 or more in a day
(or 8 or more per week) for women (NIAAA, 2005). The guide pointed out
that drinking at lower levels may be problematic depending on many factors,
such as age, co-existing conditions, and medication use.
Event-Level Data Collection
While the above methods are used to get a picture of the relationship between
an individual's overall or typical drinking behavior and risk for alcohol-related
problems or consequences, studies are sometimes interested in the amount
of "event-specific" drinking that takes place on a particular
drinking occasion. For example, researchers interested in alcohol's contribution
to violent crime or to HIV risk are interested not only in the respondent's
overall drinking habits but also in the respondent's drinking on a particular
occasion when a crime was committed or when unsafe sex occurred.
Event-level data, in which each “event” (or each day) is its own
data point, capture the complex patterns of drinking and other high risk
behaviors that typical aggregate approaches cannot
6
. They provide
more detailed data on the specific context of the drinking occasion, allowing
a researcher to better record the context for more diverse substantive
research regarding frequency, quantity, and variability in alcohol use
patterns and the co-variation between alcohol use and associated behaviors.
Because of their richness, methodologies that incorporate event-level
data are becoming more common in alcohol research.
Until recently, collection of event-level data was appealing, but practically
difficult. However, recent advances in technology have enhanced the ability
of researchers to
collect valid event-level data and analyze the data in ways that capitalize on
the complexity of the data. Event-level data may be collected in a variety
of ways each of which represents an advance beyond the collection of aggregate
data. Early efforts to collect more ecologically valid data utilized the
timeline follow-back procedure (TLFB) (Sobell and Sobell 2000), or related
measures. These methods allow for estimation of quantity of alcohol consumption,
as well as risk behaviors. Prospective methods have also been used to
assess daily alcohol use via paper-and-pencil daily drinking diaries and,
more recently through the use of advanced technology including palm-pilot
or personal data assistant (PDA) assessments, interactive voice response
(IVR), cell phones, and the internet.
RESEARCH QUESTIONS/ DIRECTIONS
Comparisons are needed of data obtained by using the QF and GF approaches with
data obtained through diary/daily recall, TLFB, IVR, and PDA approaches
both in terms of volume estimates and in terms of accuracy in capturing
overall drinking frequency and frequency of risk drinking.
- What are the advantages and disadvantages of particular event-level data collection
procedures for particular respondent sub-groups and sample sizes?
- Are current measurement techniques equally suitable for all population
subgroups (e.g. different age groups, ethnicities). Furthermore, serious
challenges remain in adapting measurement techniques to the drinking
patterns of other societies, especially tribal cultures where drinking
may be a communal activity delineated in terms of time rather than quantity.
- How well do existing measurement approaches capture atypical light drinking
among subgroups whose predominant drinking pattern is one of infrequent
heavy drinking?
B. NOSOLOGY OF ALCOHOL USE DISORDERS
METHODOLOGICAL ISSUES
“The most widely used definitions for alcohol use disorders are those determined
by editions of the Diagnostic and Statistical Manual of Mental Disorders
(DSM) of the American Psychiatric Association and the International Classification
of Diseases (ICD) of the World Health Organization. These are dichotomous
categories. Alcoholism treatment studies, human genetics studies, and
epidemiology all rely on these definitions, which constitute a near-universal
feature of research on alcoholism. Studies consistently show high reliability
for DSM-IV and ICD-9 alcohol dependence but lower reliability for alcohol
abuse/harmful use. Validity studies indicate that DSM-IV and ICD-9 alcohol
dependence diagnoses have good validity, but the validity for alcohol
abuse/harmful use is much lower. The hierarchical relationship of abuse
to dependence may contribute to the reliability and validity problems
of abuse ----.” (Hasin, 2003)
The American Psychiatric Institute for Research and Education (APIRE) convened
a panel of international experts in each of the major disorders in order
to review the evidence base for proposing revisions in diagnostic criteria
for DSM nomenclature. The results of that panel's discussions are summarized
here
7
. The process of creating the DSM-V definitions is due
to begin in 2008, and the publication of the definitions is scheduled
for 2012. The key question being addressed by the panel is whether the
DSM definitions should remain categorical or whether a dimensional component
should be included in the revision.
Bucholz (in Helzer et al., 2006) maintains that this issue extends beyond substance
abuse nomenclature to all of psychiatric classification. She argues that
one perspective that has been given little attention in the discussion
of dimensionality is that of a longitudinal perspective. Bucholz points
out that course of disorder has been an important element of diagnostic
validity, so that the extent to which a dimensional approach may enhance
that construct merits consideration as the process of revision of diagnostic
criteria begins. In order to illustrate this, she presents data from the
collaborative study on the genetics of alcoholism (COGA). Categorical
DSM-IV alcoholism dependence diagnoses were cross-tabulated at two points
in time, about five years apart. The cross-classified data suggest that
the phenotype of alcohol dependence is very unstable. However, in evaluating
the results from a latent transition analysis, the overwhelming impression
is one of stability. Bucholz suggests that findings such these may encourage
further elaboration of a dimensional sub- classification to improve both
patient care and research.
Bierut (in Helzer et al., 2006) cites evidence that alcoholism is a diagnosis
based on a continuum (e.g. Bucholz et al., 1996), with increasing severity
of illness. Using data from COGA, she investigated the question of whether
evidence of a continuum extends beyond alcohol-related conditions to comorbid
disorders and treatment. Comorbid conditions such as smoking, drug dependence
and major depressive disorder were examined in alcohol-dependent and non-dependent
individuals using an ordinal measure of lifetime DSM-IV symptoms count
compared with a categorical DSM-IV lifetime diagnosis of alcohol dependence.
The prevalence of the comorbid conditions among dependent and non-dependent
men and women using a categorical classification was greater among the
dependent groups. Furthermore, there were increasing prevalences of the
comorbid conditions correlated with the number of lifetime DSM-IV alcohol
dependence symptoms. Therefore, alcoholism was shown to lie on a severity
index. Treatment interventions were found to follow similar patterns.
Bierut concluded that the simple DSM-IV symptom count provides more information
than can be obtained with the categorical diagnosis alone.
Helzer and Guth (in Helzer et al., 2006) propose that the diagnostic work group
create the new DSM-V categorical definition as in the past. After the
categorical definition has been specified, a dimensional approach would
be created to reflect the categorical definition. They affirm that their
proposal would offer both clinicians and investigators the convenience
of a categorical diagnosis for communication, coding, and investigative
assignment, among other purposes, but it would also offer the greater
breadth of a dimensional score for measuring initial severity and tracking
treatment progress or for adding statistical power to investigative designs.
Schuckit (in Helzer et al., 2006), supports Helzer and Guth’s proposal
to base diagnoses on categories and offer dimensions as qualifiers. He
suggests that this approach might be useful across the entire DSM-V and
need not be limited to substance use disorders. Schuckit also suggests
that a combined categorical and dimensional approach would be very useful
if a companion research manual is developed to go along with the DSM.
Such a manual would inform researchers how they might best operationalize
concepts within the DSMs.
A recent study used item response theory to determine whether the DSM-IV diagnostic
criteria for alcohol abuse and dependence are arrayed along a continuum
of severity (Saha et al., 2006). Saha et al. point out that: 1) only a
few studies in the alcohol field have applied Item Response Theory (IRT)
to examine the possibility of an alcohol problem continuum in which multiple
alcohol problems map to a broad dimension of severity; 2) the samples
used in previous studies were not representative of all individuals with
alcohol problems; and 3) the sample sizes were too small to capitalize
on IRT methodology to examine whether each criterion functioned differently
among subtypes of the general population defined in terms of gender, age,
and race-ethnicity in a manner independent of the distribution of the
construct across these groups. On the basis of their results, which were
based on the 2001-2002 National Epidemiologic Survey on Alcohol and Related
Conditions, Saha et al. concluded that: 1) DSM-IV diagnostic criteria
for alcohol abuse and dependence form a continuum of severity, calling
into question the abuse-dependence distinction and the interpretation
of abuse as a milder disorder than dependence; 2) the criteria tapped
the more severe end of the alcohol use disorder continuum, highlighting
the need to identify other criteria to capture the mild to intermediate
range of severity; and 3) the drinking larger amounts or longer than intended
dependence criterion may be a bridge between drinking patterns that incur
risk of alcohol use disorder at the milder end of the continuum, with
tolerance, withdrawal, impaired control and serious social, and occupational
dysfunction at the more severe end of the continuum.
Carpenter et al. (2006) point out that the DSM-IV alcohol dependence diagnosis
includes problem drinkers with different etiological pathways to dependence,
comorbid diagnoses, and varying manifestations of alcohol problems (Grant
2000; Grant et al., 1992; Hasin and Grant, 2004). They suggest that this
heterogeneity may obscure etiological factors relevant only to specific
subgroups of problem drinkers or attenuate the predictive power of risk
factors already identified. There have been numerous attempts to identify
more homogeneous groups of dependent drinkers (Babor, 1996; Cloninger,
1987; Morey and Skinner, 1986; Zucker, 1994). Overall, multidimensional
subgrouping procedures have demonstrated more discriminating and predictive
power than unidimensional procedures that are based only on dependence
diagnosis, drinking behavior, or antecedent factors of drinking.
The Type A-Type B distinction (Babor et al., 1992) is a multidimensional model
for identifying subgroups of dependent drinkers based on a statistical
grouping procedure (k-means cluster analysis). This model has been used
in numerous studies and has been found to be applicable among alcohol
and drug dependent and non-dependent selected samples (e.g. Babor et al.,
1992; Ball, 1996; Basu et al., 2004). Type A dependence is characterized
by less familial risk, fewer physical and social consequences, milder
dependence, less psychopathology, and a more favorable course following
treatment. Greater familial alcohol problems, more severe alcohol dependence,
polysubstance use, psychiatric comorbidity, and a more chronic course
characterize Type B dependence. Carpenter et al.( ibid.) note that studies
of the Type A-Type B distinction have demonstrated discriminate and predictive
validity in treated alcohol dependent samples (Ball, 1996; Ball et al.,
2000; Morgenstern et al., 1998; Schucket et al., 1995) as well as in community
samples of dependent substance abusers (Feingold et al., 1996). Therefore,
the validity of the A- B distinction has held for both treated and untreated
dependent populations.
Carpenter et al. (2006) investigated the Type A-Type B distinction in a community
sample of problem drinkers as well as its predictive validity among problem
drinkers with no history of dependence. The results demonstrated that
the Type A-Type B dimensions have prognostic significance greater than
those of an abuse diagnosis or the number of dependence symptoms. They
maintain that by predicting differences in the onset of dependence among
problem drinkers with no history of alcohol dependence, the distinction
can be an important method for identifying a group of problem drinkers
who may account for a relatively large proportion of those with alcohol
problems.
Chung et al. (2005) point out that diagnostic criteria for alcohol use disorders
(AUDs) were derived largely from clinical and research experience with
adults, and have been applied to adolescents with no modification of the
criteria or diagnostic thresholds. Although research generally supports
the reliability and validity of diagnostic interviews with adolescents,
research also indicates that certain symptoms, particularly tolerance
and drinking more or longer than intended, may not be appropriately scaled
or operationally defined for the developmental period of adolescence (Chung
et al., 2005). Furthermore, the high yet variable prevalence of these
symptoms has had a significant impact on estimates of AUD prevalence in
teens. Multiple developmental trajectories of adolescent alcohol disorders
have been found (e.g. Schulenberg et al., 2001) and have been characterized
as developmentally limited or persistent, with problems that may be relatively
continuous or intermittent (Zucker et al., 1994). Chung et al. (2005)
conclude that a key issue for future research involves increased understanding
of the course of AUDs in the context of developmental transitions, and
other substance use and co-occurring psychopathology.
RESEARCH QUESTIONS/ DIRECTIONS
- Are abuse and dependence separate entities?
- Does abuse precede dependence more than dependence precedes abuse?
- How does amount and pattern of actual drinking relate to DSM symptoms and diagnoses?
- Should DSM definitions remain categorical or should a dimensional component
be included in the revision?
- The discriminate and predictive validity of multidimensional sub-grouping procedures,
particularly the Type A-Type B distinction should be investigated. Furthermore,
the prognostic significance of the A-B distinction compared to the dimensional
component of DSM should be investigated.
- Research should be undertaken aimed at increased understanding of the course
of adolescent AUDs in the context of developmental transitions, and
other substance use and co-occurring psychopathology.
C. GENETIC EPIDEMIOLOGY AND ALCOHOL
METHODOLOGICAL
ISSUES
New strategies need to be developed
for investigating the interactions between genes for alcoholism (and alcohol-related
diseases) and environmental factors (Rutter et al, 2001, 2006; Rutter
and Silberg, 2002). The Division of Epidemiology and Prevention research
should assist in the study and implementation of gene and environment
interaction studies in several domains. A variety of environments (environmental
candidates) have already been identified through behavior genetics research
which influence the development of alcohol use disorders and related diseases
(Bouchard and McGue, 2003). NIAAA is currently identifying likely gene
candidates which will work in tandem with these environments (and vise-versa)
to promote the development of alcohol dependence and predict responsiveness
to treatment and recovery. Exposure to an environmental pathogen, whether
behavioral or biological, may be conditional on both the genes regulating
exposure and those genetic processes that are impacted directly by the
presence of the pathogen (or triggering events such as sufficient parental
care; Reiss et al., 1995; Rutter, 2005). Further conceptualization of
the G x E interaction is necessary to move on to an appropriate measurement
strategy (Almasy, 2003; Moffitt et al., 2005)
Gene-environment (GxE) interactions were previously thought to be rare for psychiatric
illnesses such as alcoholism, but findings and models are now emerging
which begin to detail potential G x E issues in the alcohol area (Boomsma
and Martin, 2002; Rutter et al., 2006). These include findings from behavioral
genetics centers and genetics studies that focus on identifying gene candidates
for alcoholism through a variety of methods which also collect individual
and environmental information. However, while there is a high level of
curiosity about G x E interactions, and speculation about whether these
interactive factors are significant as mediators of gene expression, there
is still great uncertainty about how to proceed to best identify important
interactions, document them in large population level data sets, and intervene
to prevent significant environmental exposure to risky environments (Khoury
et al., 2005). Several practical suggestions for current and future DEPR
activities are outlined below.
Collaboration with ERGE Team
A plan of action is being developed in collaboration with the Etiology of Risk
Gene and Environment Team (ERGE) which is also developing a request to
form an advisory panel with significant representation of the area of
genetic epidemiology. The focus of the genetic epidemiology will be on
life course (longitudinal) developmental research issues and the role
of G x E interaction/interplay in early onset drinking. This initiative
will take advantage of current epidemiological data sets and those augmented
with genetic information and employ a wide diversity of transdisciplinary
approaches (Goldsmith et al., 2003)
Genetic epidemiology studies are needed to identify gene-environment interactions
with environmental factors including lifestyle (smoking, diet, physical
activity), medications, occupational exposures, and infectious diseases.
Particularly challenging is the need to develop statistical methods for
studying the multi-factorial origins of alcohol-use disorders and other
alcohol-related outcomes. Gene-environment studies can help to target
preventive efforts by identifying individuals at highest risk.
Secondary analysis of existing data can be informative on several levels. In
datasets with genotypes, direct studies of gene-environment interaction
can be undertaken. In datasets without genotypes, but with family history
of alcoholism, clues about gene-environment can be discerned. In studies
without genotypes or family history, relationships between drinking, drinking
patterns, and disease outcomes or biomarkers can identify important environmental
agents that should be considered in datasets with genotypes or family
history.
RESEARCH QUESTIONS/ DIRECTIONS
- Identify Highest Risk Individuals, Groups, and Environments: The individuals
at highest risk for alcohol use disorders as a result of the interaction
of predisposing vulnerabilities and exposure to environmental triggers
need to be identified. This requires that all components of the G x
E equation be measured accurately (Botto and Khoury, 2004). There needs
to be a parallel procedure to gene sampling for assessing a variety
of environments that may be important for the gene by environment perspective.
- Develop Conceptual and Statistical Models for Analysis of Large Longitudinal
Cohort Datasets: Promote active management and analysis of large genetic
epidemiologic datasets incorporating genetic association studies using
traditional epidemiologic measures and collection of biologic samples.
Such analyses would allow for varied designs, including prospective
cohort or case-control studies and retrospective examination of environmental
and gene by environment interaction effects (Davey Smith et al., 2005;
Moffit et al., 2005).
- Individual and Family Genetic Counseling: Develop a counseling protocol
appropriate for individuals with a high degree of familial risk. An
important prevention intervention is the implementation of counseling
for individuals at greatest risk for developing alcohol dependence.
VI. SPECIAL INITIATIVES: GRANTS* AND CONTRACTS
In the preceding sections we have overviewed key issues of concern for the Division
of Epidemiology and Prevention Research (DEPR). In this section, we briefly
discuss a number of additional areas which NIAAA has already identified
as priorities, through prior initiatives, through its research emphasis
teams, through NIH set-asides, or through contracts for specific projects.
These are discussed here so that the investment in these areas can be
understood in the context of DEPR’s entire portfolio and NIAAA’s
overall priorities.
A. COLLEGE DRINKING
*
BACKGROUND
The use of alcohol by college students is accepted by many as a rite of passage.
Drinking is considered a normal, even expected, behavior among college
students -- data from several national surveys indicate that about 80
percent of college students drink. Indeed, many freshmen come to college
with an established pattern of drinking developed in high school and even
middle school.
Since 1976, when NIAAA issued its first report on abusive drinking by college
students, research advances have transformed our understanding of alcohol
abuse and related problems. For example, we now know that a broad array
of factors affect college student drinking behavior. These include an
individual’s susceptibility to alcohol, campus norms related to
drinking, and conditions within the larger community that make alcohol
readily accessible and fail to penalize inappropriate use. Together, these
influences contribute to a culture of drinking that is more damaging and
deadly than previously recognized.
MAGNITUDE OF THE PROBLEM
Despite the widespread acceptance of drinking on college campuses, alcohol-related
tragedies never fail to create shock and generate calls for immediate
action. High-risk drinking results in unnecessary deaths, serious injuries,
assaults, and other health and academic problems. The consequences of
excessive drinking affect virtually all college campuses, college communities,
and college students, whether they choose to drink or not
.
The fact that so many people other than drinking college students experience
negative consequences from college student drinking underscores the need
for colleges and surrounding communities to learn how to more effectively
reduce these problems and act on that knowledge.
- •
Death: 1,700 college students between
the ages of 18 and 24 die each year from alcohol-related unintentional
injuries, including motor vehicle crashes (Hingson
et al., 2005). One-half of people 18 – 24 who die in
crashes involving alcohol are persons other than the drinking driver.
- • Injury: 599,000 students between the ages of 18 and 24 are unintentionally
injured under the influence of alcohol (Hingson et al., 2005).
- • Assault: More than 696,000 students between the ages of 18 and
24 are assaulted by another student who has been drinking (Hingson et
al., 2005).
- • Sexual Abuse: More than 97,000 students between the ages of 18
and 24 are victims of alcohol-related sexual assault or date rape (Hingson
et al., 2005).
- • Unsafe Sex: 400,00 students between the ages of 18 and 24 had unprotected
sex and more than 100,000 students between the ages of 18 and 24 report
having been too intoxicated to know if they consented to having sex
(Hingson et al., 2002).
- • Academic Problems: About 25 percent of college students report
academic consequences of their drinking including missing class, falling
behind, doing poorly on exams or papers, and receiving lower grades
overall (Engs et al., 1996; Presley et al., 1996a, 1996b; Wechsler et
al., 2002).
- • Health Problems/Suicide Attempts: More than 150,000 students develop
an alcohol-related health problem (Hingson et al., 2002) and between
1.2 and 1.5 percent of students indicate that they tried to commit suicide
within the past year due to drinking or drug use (Presley et al., 1998).
- • Drunk Driving: 2.8 million students between the ages of 18 and
24 drove under the influence of alcohol last year (Hingson et al., 2005).
- • Vandalism: About 11 percent of college student drinkers report
that they have damaged property while under the influence of alcohol
(Wechsler et al., 2002).
- • Property Damage: More than 25 percent of administrators from schools
with relatively low drinking levels and over 50 percent from schools
with high drinking levels say their campuses have a “moderate”
or “major” problem with alcohol-related property damage
(Wechsler et al., 1995).
- • Police Involvement: About 5 percent of 4-year college students
are involved with the police or campus security as a result of their
drinking (Wechsler et al., 2002) and an estimated 110,000 students between
the ages of 18 and 24 are arrested for an alcohol-related violation
such as public drunkenness or driving under the influence (Hingson et
al., 2002).
- • Alcohol Abuse and Dependence: 31 percent of college students met
criteria for a diagnosis of alcohol abuse and 6 percent for a diagnosis
of alcohol dependence in the past 12 months, according to questionnaire-based
self-reports about their drinking (Knight et al., 2002).
TRENDS
Several of the consequences cited above have shown changes over time. The following
are data as reported in 2002 compared with data as reported in 2005:
| |
2002 |
2005 |
| •
Deaths (per year)............................................. |
1400
|
1700 |
|
• Injuries (per year).................…………........... |
500,000
|
599,000 |
| •
Assaults (per year)………………………...... |
600,000 |
696,000
|
| •
Sexual Abuse (per year)...………………..... |
70,000 |
97,000
|
|
• Drunk Driving (per year)…………….. |
2.1 million |
2.8 million |
POPULATIONS AT RISK
The proportion of college students who drink varies depending on where they live.
Drinking rates are highest in fraternities and sororities, followed by
on-campus housing (e.g., dormitories, residence halls) (Presley et al.,
1996a, 1996b; Wechsler et al., 1998, 2000b). Students who live independently
off-campus (e.g., in apartments) drink less, while commuting students
who live with their families drink the least (O’Hare, 1990; Wechsler
et al., 2002).
A number of environmental influences working in concert with other factors may
affect students’ alcohol consumption (Presley et al., 2002). Colleges
and universities where excessive alcohol use is more likely to occur include
schools where Greek systems dominate (i.e., fraternities and sororities),
schools where athletic teams are prominent, and schools located in the
Northeast (Presley et al., 1996a, 1996b; Wechsler et al., 1996, 1997,
1998, 2000b; Werner and Greene, 1992).
Some first-year students who live on campus may be at particular risk for alcohol
misuse. During their high school years, those who go on to college tend
to drink less than their non-college-bound peers. But during the first
few years following high school, the heavy drinking rates of college students
surpass those of their non-college peers, and this rapid increase in heavy
drinking over a relatively short period of time can contribute to difficulties
with alcohol and with the college transition in general (Schulenberg et
al., 2001). Anecdotal evidence suggests that the first 6 weeks of enrollment
are critical to first-year student success. Because many students initiate
heavy drinking during these early days of college, the potential exists
for excessive alcohol consumption to interfere with successful adaptation
to campus life. The transition to college is often so difficult to negotiate
that about one-third of first-year students fail to enroll for their second
year (Upcraft, 2000).
- • Students who Drink the Most (Johnston et al., 2001; Meilman et
al., 1994, 1999; Presley et al., 1996a, 1996b; Wechsler et al., 1996,
1997, 1998, 2000b):
- o Males
o Whites
o Members of fraternities and sororities
o Athletes
o Some first-year students
• Schools Where the Least Amount of Drinking Occurs
(Meilman et al.,
1995; Presley et al., 1996a, 1996b; Wechsler et al., 200b):
- o 2-year schools
- o Religious schools
- o Commuter schools
- o Historically Black Colleges and Universities (HBCUs)
RESEARCH PROGRESS
New techniques have enabled researchers to compare alcohol-related problems in
large groups of college students and their non-college peers and to map
the extent of these problems, nationally and regionally. With this information,
researchers can determine how new laws and policies, alcohol-prevention
programs, and trends in the general population affect drinking patterns
among college students and their non-college peers.
Research shows that a number of personal factors, from family background to alcohol
use during high school, influence college students’ drinking patterns.
And as discussed previously, additional factors in the college environment
contribute to drinking patterns; for example, membership in fraternities
and sororities, sports teams, or other social groups and college organizational
factors such as size, location, and number of commuter students.
In April 2002, NIAAA issued a major report on college drinking: A Call to
Action: Changing the Culture of Drinking at U.S. Colleges. This report
was developed by the NIAAA-supported Task Force on College Drinking, a
group consisting of college presidents, researchers, students, and NIAAA
staff. The report describes new understandings of dangerous drinking behavior
by college students and the consequences for both drinkers and non-drinkers.
To provide practical assistance to colleges and universities, the Task Force
developed a series of recommendations on integrating research-based principles
and practices in alcohol program planning. The Task Force also prepared
recommendations specifically for researchers and NIAAA on the direction
of future research and areas for potential collaboration with colleges
and universities. All recommendations were based on scientific evidence
and represent the most objective guidance available on preventing risky
drinking by college students.
3-in-1 Framework
The research strongly supports the use of comprehensive, integrated programs
with multiple complimentary components that target: (1) individuals, including
at-risk or alcohol-dependent drinkers, (2) the student population as a
whole, and (3) the college and the surrounding community (Hingson and
Howland, 2002; DeJong et al., 1998; Institute of Medicine, 1989). The 3-in1 Framework
focuses simultaneously on each of the three primary
audiences and is designed to encourage consideration of multiple stakeholders,
both on and off campus.
T
ask Force Recommendations
The Task Force identified four categories (tiers 1-4) of prevention strategies.
Tier 1 strategies are those with evidence of effectiveness among
college students: combining cognitive-behavioral skills with norms clarification
and motivational enhancement interventions; offering brief motivational
enhancement interventions; and challenging alcohol expectancies. Tier
2 strategies are those with evidence of success with general populations
that could be applied to college environments: increased enforcement of
minimum drinking age laws; implementation, increased publicity, and enforcement
of other laws to reduce alcohol-impaired driving; restrictions
on alcohol retail outlet density; increased prices and excise taxes on
alcoholic beverages; responsible beverage service policies in social and
commercial settings; formation of a campus and community coalition involving
all major stakeholders may be critical to implement these strategies effectively.
Tier 3 strategies are those with evidence of logical and theoretical
promise, but require more comprehensive evaluation: adopting campus-based
policies and practices that appear to be capable of reducing high-risk
alcohol use; increasing enforcement at campus-based events that promote
excessive drinking; increasing publicity about and enforcement of underage
drinking laws on campus and eliminating “mixed messages;”
consistently enforcing disciplinary actions associated with policy violations;
conducting marketing campaigns to correct student misperceptions about
alcohol use; provision of “safe ride” programs; regulation
of happy hours and sales; informing new students and their parents about
alcohol policies and penalties before arrival and during orientation periods.
Tier 4 strategies are those with evidence of ineffectiveness:
informational, knowledge-based, or values-clarification interventions
about alcohol and the problems related to its excessive use, when used
alone; providing blood alcohol content feedback to students.
Since the 2002 NIAAA report, several additional studies have reported positive
benefits of individually-oriented screening and brief intervention (Larimer
et al, in review). Two studies have identified reductions in drinking
and alcohol-related problems specifically among college students associated
with the implementation of comprehensive college/community programs focused
on environmental strategies to reduce alcohol availability and violation
of drinking driving statutes (Weitzman et al, 2004; Clapp et al, 2005).
Rapid Response to College Drinking Problems
In December 2002, NIAAA issued an RFA AA-03-008: “Research Partnership
Awards for Rapid Response to College Drinking Problems.” The purpose
of this RFA was to provide a mechanism for developing and conducting research
requiring rapid funding to address unexpected and time-sensitive research
opportunities for college campus drinking projects. This RFA was aimed
at established alcohol research scientists with expertise in college drinking
who could serve as resources for college and university administrators
experiencing alcohol-related problems on their campuses. Five (5) U01
grants were awarded in September 2003. Each grant is for up to 5 years
and provides up to $500,000 per year in direct costs.
In June 2003, NIAAA issued a companion solicitation: PAR-03-133; “Rapid
Response to College Drinking Problems.” This PAR was intended to
support studies of services or interventions that could capitalize on
“natural experiments” (e.g., unanticipated adverse events,
policy changes, new media campaigns, formation of campus-community coalitions,
etc.). The intended applicants for this PAR were college administrators
who saw an urgent need to quickly address an alcohol-related problem on
their campuses. Fifteen (15) U18 grants were awarded between September
2004 and September 2005. Each grant is for up to 3 years and provides
up to $200,000 per year in direct costs.
The key feature of the rapid response program is that each U18 college grantee
is required to partner with a U01 researcher grantee. Together, these
pairs, working with NIAAA Scientific Staff Collaborators, jointly design,
develop, implement, and evaluate college drinking projects on the U18
campuses.
These projects include individually-focused strategies and environmentally-focused
strategies. For example: protective influence of Residential Learning
Communities, peer-facilitated alcohol interventions, campus-community
coalitions, peer-led motivational enhancement with freshman women, freshman
parent-student initiative, fraternity and sorority interventions, alcohol
screening and interventions in a college health clinic, reducing high-risk
drinking among first-year students, a University Assistance Program, intervention
for freshman violators, social norms program, and reducing high-risk drinking
among students celebrating their 21st
birthday.
In addition to the 5 UO1 grants and the 15 Rapid Response grants, NIAAA is supporting
32 other grants examining college drinking and prevention. These are administered
by DEPR.
RESEARCH QUESTIONS/ DIRECTIONS
- • How does alcohol use in college influence academic trajectories? What
are the effects of different drinking patterns on test performance?
What is the effect of excessive alcohol consumption on the developing
brain?
- • What are the barriers to screening and counseling among college and underage
individuals? How can screening and brief intervention be expanded to
reach a majority of college students with alcohol use disorders and
be more effective in reducing those problems? What is the impact of
programs to reduce these barriers on students who need alcohol counseling?
Will this yield population level reductions in binge drinking and alcohol-related
problems?
- • Are interventions that combine environmental prevention initiatives with
efforts to increase screening, brief interventions, and treatment more
effective in reducing college drinking problems than programs that focus
only on a single approach?
- • What are the most effective environmental prevention initiatives to reduce
college drinking problems? (e.g. substance-free dorms, parental notification
of alcohol problems, reinstatement of Friday classes, alcohol advertising
bans on college property and in publications, and amnesty for reporting
alcohol problems).
- • How can students who are concerned about college drinking problems at
their universities be identified and given an opportunity to help plan
and implement strategies to reduce these problems, including the second-hand
effects of drinking? How effective are these strategies in reducing
alcohol-related problems?
- • How can parents be most effectively enlisted in efforts to reduce college
drinking problems?
- • Can consortia of schools be effective in addressing college drinking problems?
Is there value in identifying strategies that can be effective across
multiple settings? Will findings from groups of schools have an impact
on changing other schools’ policies and practices?
B. UNDERAGE DRINKING*
A SERIOUS AND WIDESPREAD PROBLEM
Alcohol is the drug of choice among youth. Data from the 2005 Monitoring the
Future (MTF), an annual survey of U.S. youth, show that more than three-fourths
of 12th graders, nearly two-thirds of 10th graders, and more than two-fifths
of 8th graders have consumed alcohol at some point in their lives. And
when youth drink, they tend to drink heavily. Underage drinkers consume
on average four to five drinks per occasion about five times a month.
By comparison, adult drinkers ages 26 and older consume on average two
to three drinks per occasion about nine times a month (Figure 9).
Figure 9
Adolescents Drink Less Frequently Than Adults, but Drink More Per Occasion
Number of Drinking Days per Month and Usual Number of Drinks per Occasion
for Youth (12-20), Young Adults (21-25) and Adults (26 and older).
Source: U.S. Substance Abuse and Mental Health Services Administration.
2004 National Survey on Drug Use and Health (NSDUH)
A particularly worrisome aspect of underage drinking is the high prevalence of
heavy episodic drinking, defined as drinking five or more drinks in a
row (or five for boys and four for girls depending on the survey) in the
past two weeks (or thirty days depending on the survey). MTF data show
that more than 10 percent of 8th graders, 20 percent of 10th graders,
and 25 percent of 12th graders engage in heavy episodic drinking. Underage
drinking can result in a range of adverse short-and long-term consequences,
including academic and/or social problems; physical problems such as hangovers
or illnesses; unwanted, unintended, and unprotected sexual activity; physical
and sexual assault; memory problems; increased risk of suicide and homicide;
alcohol-related car crashes and other unintentional injuries such as burns,
falls, and drownings; and death from alcohol poisoning (NIAAA, 2004/2005).
Recognizing the seriousness of the problem, NIAAA has designated underage drinking
as a research priority area. In addition, the convergence of a number
of developments contributes to the timeliness of this research focus.
First, new findings from NIAAA-supported research indicate that the kind
of serious drinking problems previously associated with middle adulthood
(including what has been called alcoholism) often begin to emerge during
adolescence and young adulthood (Figure 10).
Figure 10
18-24 Year Olds Have the Highest
Prevalence of DSM-IV Alcohol Dependence
Prevalence of Past-Year DSM-IV Alcohol Dependence – U.S.
Source: National Epidemiologic Survey on Alcohol and Related Conditions
(NESARC) 2001-2002
These findings, and other results that enhanced our understanding
of alcohol consumption during adolescence, have led to a reconceptualization
of alcohol dependence within a developmental framework. Second, accumulating
neurobiological research suggests that adolescence may be a period of
particular vulnerability to the effects of alcohol. Third, the National
Research Council and Institute of Medicine report, Reducing Underage
Drinking: A Collective Responsibility, emphasized the dangers of underage
drinking, even when the level of drinking falls short of a diagnosable
condition. And finally, Congress established the Interagency Coordinating
Committee on Preventing Underage Drinking (ICCPUD) further underscoring
the importance of the problem. NIAAA provides the scientific foundation
for ICCPUD.
UNDERAGE DRINKING AND DEVELOPMENT
Pervasive drinking by youth and the emergence of alcohol misuse and dependence
in late adolescence, are intertwined with developmental processes. Significant
changes occur in the body over the course of adolescence, including hormonal
alterations, and the formation and pruning of neural networks in the brain.
Adolescence is also a time of heightened emotions, increased risk taking,
and trying new experiences and activities that emphasize socializing with
peers and conforming to peer-group standards. This confluence of factors
may place young people at particular risk for initiating and escalating
alcohol consumption. The overarching goal of NIAAA’s initiative
on underage drinking is to enhance understanding of alcohol-related problems
by approaching them in a developmental framework and to intensify research,
evaluation, and outreach efforts to address this important public health
problem which affects our Nation’s young people.
THE WORK OF THE TEAM AND ITS ADVISORS
About two years ago the Underage Drinking Team recruited a Steering Committee
of expert advisors in order to help it accomplish its goals. This interdisciplinary
group of advisors includes experts on adolescent alcohol use and problems,
on overall adolescent development, and on policy and communications. The
Steering Committee has met with the Underage Team three times over the
past two years to advise the team as it formulates its research priorities.
The priorities that emerged include: a better understanding of adolescent
decision making, the effect of alcohol on the developing brain, and the
special case of very early initiators.
To date, the team has published a special issue of Alcohol Research and Health
entitled “Alcohol and Development in Youth: A Multidisciplinary
Overview,” issued an RFA entitled “Underage Drinking: Enhancing
Healthcare System Responses for which the awards will be made this summer,
and participated in a NIDA RFA addressing substance use and the developing
brain (two NIAAA awards were made). In addition, working groups of the
Steering Committee have produced drafts of three age specific (under 10,
10-15 and 16-20) reports which will form the basis of a developmentally
focused report on underage drinking. The work of the team has also provided
the scientific foundation for the upcoming Surgeon General’s Call
to Action on Underage Drinking. For further information about the Initiative
please go to the following web address:
http://www.niaaa.nih.gov/AboutNIAAA/NIAAASponsoredPrograms/underage.htm .
UNDERAGE DRINKING AND THE DIVISION OF EPIDEMIOLOGY AND PREVENTION RESEARCH
The work of the Underage Team and the Division are closely linked. Many
of the studies funded by NIAAA that focus on underage alcohol consumption,
in particular, epidemiologic studies of risk and protection, and prevention
trials, are in the portfolio of this Division.
C. AIDS*
OVERVIEW OF HIV/AIDS PREVENTION ACTIVITIES FOR NIH
“The human immunodeficiency virus (HIV), the cause of AIDS, has now infected
more than 65 million people worldwide, of whom 25 million have died, according
to estimates of the Joint United Nations Programme on HIV/AIDS (UNAIDS).
AIDS is the deadliest pandemic of our generation and one of the worst
in history. In the year 2005 alone, approximately 4.1 million people worldwide,
half of them women, became infected with HIV, and about 2.8 million individuals
with HIV/AIDS died. More than 95 percent of these infections and deaths
occurred in developing countries, most of which are also burdened by other
significant health challenges. In these nations, HIV/AIDS threatens not
only human welfare, but social, political and economic stability as well.
In the United States, an estimated one million people are living with HIV infection,
and 40,000 new infections occur each year. HIV infection rates continue
to climb among women, racial and ethnic minorities, young homosexual men,
individuals with certain addictive disorders and people over 50 years
of age. To date, HIV/AIDS has killed more than half a million people in
our country.
The magnitude of the HIV/AIDS pandemic requires a robust, multi-faceted and sustained
global response from all sectors of society. At the U.S. National Institutes
of Health (NIH), we are committed to accelerating the research that will
help end the scourge of HIV/AIDS. The NIH effort represents the largest
public investment in HIV/AIDS research anywhere in the world. In fiscal
year 2006, $2.9 billion has been appropriated for NIH AIDS-related research.
The resources devoted to HIV/AIDS research by our many partners in the
research enterprise also have been significant.”
Statement of Elias A. Zerhouni, M.D., Jack Whitescarver, Ph.D., and Anthony S.
Fauci, M.D., National Institutes of Health on The 25th Anniversary of
the First Published Reports of AIDS (http://www.nih.gov/news/pr/jun2006/niaid-01.htm)
Alcohol and HIV/AIDS
To respond to the AIDS pandemic, the NIAAA in conjunction with NIH has developed
a biomedical and behavioral research program to better understand the
basic biology of HIV, develop effective therapies to treat and control
HIV disease, and design interventions to prevent new infections from occurring
in alcohol using, abusing and dependent individuals. The FY 2006 and FY2007
research agenda continues the following over-arching themes: research
to prevent and reduce HIV transmission; research to test the impact of
therapies for those who are already infected; international research,
particularly to address the pandemic in developing countries; and biomedical
and behavioral research targeting the disproportionate impact of AIDS
on minority populations in the United States.
SCOPE OF RESEARCH WITHIN DEPR: A SUMMARY
The alcohol and AIDS area has a separate but overlapping mandate with the Division
of Epidemiology and Prevention Research within NIAAA. The prevention and
treatment of HIV/AIDS includes the study of the epidemiology of risk factors
including alcohol use, abuse, and dependence which contribute to HIV infection,
disease manifestation, clinical decision making, and interventions to
prevent transmission. These areas will be only briefly described here
because of their separate research objectives generally outlined in the
NIH Plan for HIV/AIDS Research, their separate funding status, and plans
for separate Extramural Advisory Board review activity in the future.
Currently NIAAA funds approximately 60 grants broken down within the areas
of Epidemiology (11), Primary Prevention (29) and Basic Behavioral Research
for HIV/AIDS (21). Each of these areas represents a substantive focus
for alcohol and AIDS activities at NIAAA with cross-cutting implications
for special populations of minorities, youth, women, gay men, and other
special populations.
INDEPENDENT EXTRAMURAL ADVISORY BOARD FOR HIV/AIDS
An Extramural Advisory Board meeting to discuss strengthening Alcohol and AIDS
biomedical research is planned for October. This meeting will not directly
address behavioral prevention issues present in the applications within
DEPR. It is expected that an additional meeting(s) focusing on Alcohol
and AIDS Prevention Science Initiatives will be targeted for FY2007 and
FY2008.
EPIDEMIOLOGICAL GOALS
Surveys and qualitative studies of the general population provide information
on levels of infection among alcohol users and their partners, culturally-related
risk factors for HIV transmission, and information on appropriate venues
in which to implement alcohol and HIV prevention strategies. In addition,
clinical epidemiological samples of AIDS patients contribute greatly to
our knowledge of how HIV+ individuals with alcohol use disorders manifest
the disease-related consequences of chronic and acute infection. These
epidemiological studies inform prevention researchers and clinicians about
screening, diagnosis, and treatment of HIV+ individuals. Two clinical
cohorts of HIV+ alcohol dependent individuals and matched case controls
which support alcohol and AIDS research are of note, The Veteran’s
Aging Cohort Study in collaboration with the Veteran’s Administration
(VACS) and the HIVLIVE, an earlier clinical cohort based in the Boston
area. Both have been productive in identifying both behavioral and biomedical
issues for these populations in the context of the changing epidemic and
treatment availability.
NIAAA supports approximately 11 grants in the combined area of Natural History
and Epidemiology for a total cost of 1.2 million dollars.
PREVENTION GOALS
Behavioral prevention issues for alcohol and AIDS research are also represented
within the portfolio of DEPR. These prevention issues are divided primarily
into two areas. First those applications which examine how levels of alcohol
use and the psychological and social processes impacted by alcohol use
make individuals more likely place themselves at risk for HIV infection
through sexual risk taking and to a lesser extent by combining alcohol
and IV drugs. This focus informs primary prevention activities through
schools, educational programs, bars, peer influence, modifications of
drinking environments, server training and safer sex messages, and general
public health campaigns targeted at high-risk communities. New interventions
are being proposed which focus on social and structural factors which
have been successfully used to avert or reduce levels of infection in
specific targeted populations particularly international populations (military,
sex workers, etc). These Alcohol and AIDS prevention activities which
have proved to be successful are being translated through the Presidential
Emergency Plan for AIDS Relief (PEPFAR) into broadly disseminated interventions.
NIAAA supports approximately 29 grants in the area of Preventive Interventions
for both domestic and international populations with a special emphasis
on women and minorities for a total of 10.7 million dollars. In addition,
21 grants are supported in the Basic Behavioral Sciences area for 4.5
million dollars.
There are three distinct alcohol and HIV/AIDS prevention goals:
Primary Prevention
The first goal is to develop preventive interventions targeted at “at-risk”
populations and consist of primary prevention for uninfected individuals
who misuse alcohol either as a result of hazardous or harmful consumption
or as a result of participation in high risk social networks or venues
such as bars. These programs focus on adolescents, youth, women, gay men
and other special populations impacted by drinking and risk for HIV. Interventions
include brief motivational counseling, testing and screening, and cognitive
behavioral, group, social, and structural interventions through changes
in policies. For example, understanding the role of bar-based culture
for gay men was critical in the development of effective interventions
by “opinion leader.” These interventions are based on the
use of social diffusion of innovation theory and provide public health
messages and interventions focused on safer sex and condom distribution.
However, these interventions do not directly take into account the impact
of past and current alcohol consumption on judgment and decision-making
processes. Improved interventions are being proposed to directly address
the effects of alcohol.
Basic Behavioral Science
The second focus is on developing the theoretical underpinnings for intervention
research through the identification of risk factors and risk processes
that may occur at the individual, group or environmental - structural
level contexts. Theoretical frameworks are evolving to take into account
alcohol-specific theories such as cognitive narrowing, “alcohol
myopia”, decisional balance, sexual expectancy, and physiological
arousal. Future workshops and conferences are planned to increase this
“theory to practice” relationship and increase intervention
effectiveness. Further development and pairing of theory-based interventions
with specific settings, situations, and multilevel interventions will
increase the benefits of alcohol-specific theories.
Preventing Negative Consequences of Infection
The third focus is on preventing the negative consequences of HIV infection and
AIDS treatment. AIDS is now being treated as a chronic disease in many
countries where anti-retrovirals are readily available. However, stigmatization
continues to have a negative impact on the access and availability of
treatment, and subsequently the quality of patients’ life. Interventions
have been characterized by screening for HIV and early intervention to
change the course of the disease. In general, alcohol users who are HIV+
are not easily recruited or retained in interventions and often show up
later in the course of their disease (with more medical complications)
than those without alcohol problems. The avoidance of medical interventions
for their HIV has multiple consequences from the point of view of treatment
and coinfection. Subsequently, secondary prevention for both non-adherence
to medication regimens and resulting disease progression, and sexual risk
behavior that often escalates around relapse to alcohol use can be targeted
by both behavioral and pharmacological interventions. The difficulty of
addressing multiple needs in these populations (which often includes homelessness
and psychiatric comorbidity) and the need for alcoholism treatment and
relapse prevention is quite evident.
Addressing DEPR Themes: Four areas of complimentary research:
1) Measurement issues: Event Level and Social Outcomes
Measurement of alcohol consumption is important in establishing past and present
host response to potential infection events. This requires not only the
accurate measurement of alcohol quantity and frequency (e.g. where the
alcohol is “home brew” and of unknown alcohol content as is
often encountered in international settings), but also the proximity of
the alcohol use to a specific sexual or needle use activity that results
in infection. It remains important to understand the causal links between
risk behavior and infection events that are mediated by cognitive processes
such as “alcohol myopia” or physiological disinhibition. These
relationships also extend to biological processes such as impaired host
immunity or viral shedding, which may impact susceptibility and infectivity
respectively.
These broader dimensions around disease transmission need to be integrated into
models that estimate total population costs (and benefits) for differential
rates of transmission of disease and disease progression among alcohol
users. As therapeutics become more effective among individuals without
alcohol use disorders, the potential benefits increase for these individuals.
These individuals may live up to an additional 30 years after infection
with adequate treatment. However, recent research has shown that the direct
effects of hazardous drinking may reduce life expectancy by 6-12 years
and that any drinking may be risky and result in treatment failure and
premature mortality (Braithwaite et al., 2005).
2) Nutrition among HIV+ Alcohol Abusing Pregnant Women
Nutritional status remains an important mediator of HIV treatment outcome. In
most developing countries where anti-retrovirals (ARVs) are being offered
an additional nutritional package of vitamins (micronutrients) and food
are being offered to boost the effectiveness of ARVs. Research on nutrition
among alcohol users, and potentially HIV+ alcohol users, indicates that
there may be specific nutritional deficits and organ damage (liver) that
needs to be ameliorated prior to delivery of a full course of potentially
hepatotoxic ARVs. A specific subset of these impacted HIV+ Alcohol dependent
group, pregnant women, may need special attention relative to short-course
niverapine interventions for prevention of Mother to Child Transmission.
FAS research is being expanded in South Africa to begin to address this
need to more fully understand maternal and fetal outcomes in this targeted
group.
3) Structural Interventions
The need for interventions targeted at levels beyond the individual have clearly
been recognized. In the past these interventions have examined dyadic
(partner) and group level (social network) domains. High social, ecological,
and economic levels of influence need to be examined and preventive interventions
developed to impact these structural characteristics. Research clearly
shows the importance of access and availability of alcohol as a contributing
factor to the spread of sexually transmitted infections (STIs). Changes
in density of liquor outlets and taxation impact the rates of gonorrhea
in the US. Increasing density leads to higher rates while increasing taxation
and reduction in availability of inexpensive alcohol leads to reduced
rates of infection. Similar findings are expected for HIV particularly
in resource poor settings with limited regulation of alcohol sales and
distribution. This research calls for multilevel interventions to moderate
the impact of alcohol on the spread of STIs including HIV with a particular
focus on venue-based interventions. In addition, policies need to be developed
and enforced for the sale of alcohol in regions with regard to both cultural
and socioeconomic impact. The Special Presidential Initiative Outlined
below is one such preventive intervention approach being put into place
as a translational research activity in specific regions of Africa.
4) Special Presidential Prevention Initiative
Interventions to Reduce Alcohol-Related Sexually Transmitted
HIV Infection in High Risk Venues in Africa for consideration under the
Special Initiative on Alcohol, Presidential Emergency Plan for AIDS Relief
(PEPFAR)
This preventive intervention focus is currently being coordinated with the President’s
Emergency Plan for AIDS Relief to Africa. This is the extension meeting
in Tanzania, Dar es Salaam (August 2005) with the White House Office of
the Global AIDS Coordinator (OGAC); 2) DHHS: a)CDC; b) SAMHSA; 3) USAID;
4) DoD in which NIAAA (Kendall J. Bryant, Ph.D. and Deidra Roach, M.D.)
provided technical assistance for the development of FY 06 Country Operational
Plan and Country Assistance Plans to African nations that receive funds
under the President's Emergency Plan for AIDS Relief (PEPFAR). Specifically,
the technical assistance addressed the rationale and recommended approaches
to incorporate alcohol-related HIV risk reduction strategies in these
plans. The alcohol and AIDS plan is being developed further with support
from the OGAC.
Proposed Program Design and Implementation. High-risk venues provide opportunities
for intervening with a range of public health, evidence-based, low-cost
feasible interventions that can reach high risk populations and enable
them to reduce their risk taking behaviors and potential for transmission
of HIV.
This program will identify multiple sites in Africa where heavy alcohol use has
been documented in conjunction with HIV prevalence. The program will focus
on venue-based interventions that can be implemented across a variety
of sites and among appropriate target populations. Money for this activity
has been set aside by the Global AIDS coordinator to be integrated into
country operational plans for targeted sites (approximately $3 million,
FY 2007).
Implementation of this program in each site will include the following:
1) Identification of specifically targeted high risk populations (e.g. transport
workers, commercial sex workers, deployed military personnel) who engage
in alcohol-related sexual risk behaviors that increase their risk of acquiring
HIV (e.g. multiple partners, concurrent sexual partners, failure to use
condoms correctly and consistently).
2) Identification, description and mapping of high risk venues (bars, shabeens,
or other formal or informal drinking establishments) suitable for implementation
of interventions. The general context, social dynamics, and timing of
alcohol consumption and HIV risk behaviors among the population should
be described and understood.
3) Engagement and involvement, as appropriate, of key stakeholders (e.g. NGOs,
local health and social service providers, community leaders) in the community
with an interest in reducing alcohol-related, sexually transmitted HIV.
4) Training and education on alcohol-related HIV risk reduction for bar-based
personnel, including bar owners and managers, servers such as bartenders
or “beer girls,” and other related personnel, such as brothel
managers or procurers for commercial sex workers, who interact with the
target populations and are in a position to influence drinking and sexual
behaviors.
5) Development and delivery of appropriately tailored interventions for the identified
target populations. Interventions should include messages regarding alcohol
and the potential for increased HIV risk, appropriate HIV prevention messages
and skills-based training for behavioral risk reduction, and distribution
of condoms in high risk settings.
6) Depending upon the local cultural context and capacity for delivering such
interventions, a range of different models may be appropriate. Interventions
may include KIE (knowledge, information, education) or social marketing
of condom campaigns as well as group or community-level social network
or diffusion of innovation behavioral interventions.
Feasibility. Feasibility of this intervention program is high if sites
are successful in recruiting and engaging those members of the community
who profit from the sale and consumption of alcohol, namely bar owners
and personnel. Sites should also have the organizational capacity to add
the alcohol component to existing HIV prevention program activities. Each
program should be a PEPFAR funded program. The focus should be on interventions
that address reduction of HIV-related risk behaviors that take place within
the context of drinking. Interventions should also address the issue of
alcohol as it relates to adherence to treatment for HIV. Interventions
for HIV requiring the support of owners and managers have been carried
out successfully in high risk venues in the US; for example, in bars and
bathhouses frequented by males seeking males (MSM), and in brothels and
drinking establishments in Southeast Asia.
Impact. Outcome measures can be behavioral in terms of movement to action,
attitude change, or service utilization (getting tested for HIV, STDS).
Measures can also include specific alcohol-related outcomes such as unprotected
sexual intercourse with casual partners, negotiated safety, and the proportion
who knew their own and their partner's HIV status. Biomarkers such as
measures of STIs can also be used.
D. ALCOHOL POLICY INFORMATION SYSTEM
The Alcohol Policy Information System (APIS) contract (N01AA12009) supports NIAAA
by encouraging and facilitating research on the effects and effectiveness
of alcohol-related public policies by providing researchers with:
- • authoritative, detailed, and comparable information on alcohol-related
policies, including the dates on which they became effective;
- • user-searchable access to policy information through a public web site;
and
• information at several levels of detail, including charts, maps, summaries
and full text of laws
The contract was awarded to the CDM Group, Inc. in September 2001 with effective
dates of 14 September 2001 to 13 September 2006. The budget for FY 2005
was $1,491,595.
A separate advisory committee of public policy researchers and academic legal
experts has been formed to provide oversight, evaluation and advice to
NIAAA regarding APIS and future plans after September 2006. Particular
focus will be placed on stimulating expansion of use of the system by
alcohol policy researchers.
The main product of the APIS project is information on alcohol-related public
policies provided through a public web site. The APIS web site provides
user-searchable access to alcohol-related policy information at several
levels of detail (up to and including the full texts of laws and regulations).
Specific policy topics to be covered through APIS were selected on the
basis of: 1) public health significance; 2) research salience; 3) recent
or anticipated policy-making activity; 4) relatively straightforward legal
research requirements for adequate characterization; and 5) diversity
in terms of the topic areas addressed.
The contractor’s responsibilities include the following major tasks, each
incorporating rigorous quality assurance procedures to ensure the utility,
reliability, and integrity of the system and its various components:
- (1) performing legal research and related tasks to identify and acquire accurate
and detailed information documenting the status of alcohol-related public
policies adopted at the Federal and State levels and changes to those
policies (including the dates on which policy changes became effective);
- (2) developing summary indicators, measures, and variables to characterize alcohol-related
policies in ways that facilitate comparisons of policies across jurisdictions
and/or over time;
- (3) establishing and managing appropriate electronic data storage systems for
maintaining APIS policy information (including all related descriptions
and variables and relevant portions of the full-text of statutes and
regulations) to facilitate search and retrieval of policy information
by a variety of flexible criteria as required to operate and maintain
the APIS;
(4) designing, developing, operating, maintaining, and updating of the APIS web
site to provide users with searchable access to APIS policy information
http://alcoholpolicy.niaaa.nih.gov;
- (5) devising suitable display formats to represent policy characteristics and
variation through the APIS web site; and
- (6) responding to inquiries and providing technical assistance to users of APIS
information.
As a unique research resource, the APIS has been developed “from scratch”
under the present contract, with no closely analogous model available.
Plans for establishing a successor contract are currently underway.
APIS Policy Topics – The centerpiece of APIS’s services
for researchers is the detailed information provided for each of the 36
APIS policy topics. A list of current APIS policy topics has been shown
previously in Table 3 (page 68 in the Alcohol Policy section).
For each APIS policy topic, structured information is provided
in the following categories:
Policy Description – a brief narrative description of the
features and importance of the policy topic
Definitions
– where needed, definitions for specialized terms used in the description
and documentation of the policy topic
Explanatory Notes and Limitations – detailed notes identifying
specific issues that affect the documentation and interpretation of policy
information (the inclusion of these notes is one feature that distinguishes
APIS policy information from most other sources of information on specific
policies)
Variables
– definitions for the specific variables (i.e., policy attributes) used to
characterize the status of a State’s policy in the area defined by the
policy topic
Federal Law – a brief narrative summary of Federal laws pertaining
to the policy topic, followed by excerpts of relevant sections of statutory
and regulatory codes
Selected References – a short list of relevant Federal publications
pertaining to the policy topic.
Comparison Tables
– these tables provide the detailed policy
information that is of greatest use to researchers. The layout of each
type of comparison table is consistent across all policy topics. Each
set of comparison tables presents three alternative views:
- • Policies in Effect on a Specific Date: This view allows the user to specify
any date within the APIS coverage period (1/1/1998 – 1/1/2006
for most policy topics) and compare the status of policy characteristics
across States as of that date. User controls allow selection of a subset
of specific States and sorting of results according to any of the policy
characteristics. Links within the table provide the full text of relevant
sections of State statutory and regulatory codes to document the basis
for APIS coding; additional links provide explanatory notes that pertain
to individual State circumstances as needed. Data from this table view
may be downloaded in a format suitable for use by most statistical software.
- • Policy Changes Over Time: For each State (or, in the default display,
for each State in which there have been changes in policy status during
the APIS coverage period), this view displays “before-and-after”
coding of the policy status with identification of the exact dates on
which policy changes took effect. Overall layout is identical to the
Policies In Effect on a Specific Date table, except that the date column
displays date ranges instead of individual dates. Data from this table
view may be downloaded in a format suitable for use by most statistical
software.
- • Timeline of Policy Changes: This alternate view that makes identifying
the timing and general categories of policy changes across States easier
to discern than in the Policy Changes Over Time view. Table entries
in the Timeline view provide pop-up summary information about the change
in an individual State and are clickable to provide a restricted view
of the Policy Changes Over Time table for the specific State in question
(which provides full documentation of before-and-after policy characteristics
and links to full text excerpts)
Maps and Charts – Summary maps display variation in key policy
attributes across States as of the latest update of APIS information for
the policy topic. Summary charts display changes over time in the numbers
of States with particular policy attributes. The colorful figure provides
easy overviews of variation in a specific policy area in both cross-sectional
and longitudinal terms. These visuals are designed to generate interest
and draw users into the more detailed information in the comparison tables
on which the maps and charts are based. Figure 11 shows a sample map and
chart as they appear on the APIS web site.
Related Policy Topics – links to other APIS Policy Topics
that bear a significant relationship to the policy topic on display
.
Figure 11
Examples of APIS Maps and Charts
Source:
http://www.alcoholpolicy.niaaa.nih.gov/
Additional APIS Resources – APIS also provides detailed information
on alcohol-related public policies in several other categories.
- • A section on Enforcement and Compliance provides description and
documentation of data resources that can assist researchers in measuring
the policy variables, as well as scholarly literature reviews on enforcement
and compliance from the perspectives of Public Health and Public Policy
research, Sociology and Criminology, and Economics.
- •
The Enacted Bills and Adopted Regulations section of the APIS web
site provides a search tool for identifying newly-enacted policies relevant
to alcohol and links to the statutory and regulatory code sections that
they amend. Search capabilities include structured category-based search
using the Alcohol Policy Classification System, a 77-category taxonomy
with additional 35 cross-cutting categories used for classifying alcohol-related
policies, as well as full-text searching on the bills and regulations
using Boolean search strings. Work on this section was suspended for
budgetary reasons, so coverage is limited to bills enacted and regulations
adopted in 2002 and bills enacted in 2003.
• The section of the APIS web site titled NHTSA Alcohol-Highway Safety
Digest Topics provides information transcribed from the 1988-2003
editions of this hard-copy resource in several policy topic areas, including
underage drinking and BAC limits (extending APIS historical coverage
by an additional ten years), and drinking and driving (providing coverage
of topics not yet available through standard APIS procedures). The inclusion
of this material marks first time that this information has been made
publicly available in electronic format.
- • The new State Profiles of Underage Drinking Laws
section provides
a useful alternate view of eleven policy topics pertaining to underage
drinking and access to alcohol for one user-selected State at a time.
This facility of the web site was developed through discussions with
Federal Trade Commission staff who identified this summary narrative
approach to displaying information from the policy database as a useful
adjunct to a web site they were developing on the issue of underage
drinking. This feature also serves as a pilot test for possible expansion
of a State Profiles view for the larger set of policy topics.
The information on alcohol-related public policies provided through APIS is designed
to provide researchers with reliable information that can be incorporated
in analyses of the effects and effectiveness of such policy measures.
In most cases, such studies will use APIS policy information in conjunction
with data on alcohol-related behaviors and outcomes from other sources.
Examples of outcome data that could be used in conjunction with APIS policy
information in analytic studies include data from the Fatality Analysis
Reporting System on alcohol-involved traffic crashes or fatalities; measures
of average alcohol consumption by State such as those published annually
by the Alcohol Epidemiology Data System; health outcome data such as hospital
discharge statistics available from the Agency for Healthcare Quality
and Research; and survey data reflecting alcohol-related behaviors and
outcomes, such as the National Epidemiologic Survey on Alcohol and Related
Conditions and the National Survey of Drug Use and Health. Studies can
use statistical methods that capitalize on both cross-sectional and longitudinal
variation in alcohol-related policies and outcomes to identify the effects
of policy changes. In areas where there have been significant changes
in policy status over time (for example, there have been more than 40
changes in laws pertaining to underage drinking and access to alcohol
over the period from 1998 through January 1, 2006), this variation can
provide the statistical power needed to identify key policy attributes
that are most effective in reducing the adverse consequences of alcohol
consumption.
E. ALCOHOL EPIDEMIOLOGY DATA SYSTEM
Alcohol Epidemiology Data System (AEDS) Contract
The AEDS contract (N01AA32007) is a 5-year cost-reimbursement contract with a
performance-based award fee provision. The successful proposal from CSR,
Incorporated, submitted in April 2003 pursuant to RFP number AA0303, was
awarded in September 2003 with effective dates of September 16, 2003 to
September 15, 2008.
The purpose of the contract is to support NIAAA by:
- • Maintaining and acquiring alcohol-related national data sets
- • Conducting research studies utilizing alcohol epidemiologic data
• Providing analytic
reports annually on liver cirrhosis mortality, alcohol-related fatal and
other traffic accidents, and other subjects
• Responding to alcohol epidemiologic data questions
• Establishing and maintaining epidemiologic research
consultants.
The contract supports the NIAAA mission by:
- •
Conducting and supporting research in a wide range of scientific areas
related to alcohol epidemiology
• Coordinating with international, national, state, and local institutions,
organizations, agencies and programs engaged in alcohol-related work
• Translating and disseminating research findings to health care providers,
researchers, policymakers, and the public.
Specific requirements of the contract include the following 5 tasks:
Task 1: Maintenance and extension of an alcohol-related epidemiologic data bank.
This task includes acquisition of major datasets, collection of state-level and
manufacturer data on alcohol consumption, and yearly production of an
Alcohol Epidemiologic Data Directory.
The contractor continually downloads the most recent versions of major publicly
available datasets including alcohol such as the National Health Interview
Survey and the Behavioral Risk Factor Surveillance System. In addition,
the contractor continually acquires state-level data on alcohol consumption
which are included in an annual report on per capita alcohol consumption
in the US.
The Alcohol Epidemiologic Data Directory is a current listing of surveys and
other relevant data suitable for epidemiologic research on alcohol. It
is available on the NIAAA website at
http://pubs.niaaa.nih.gov/publications/surveillance.htm. Some surveys included in the Directory
are designed specifically to answer alcohol-related questions. Other surveys
may address other issues but still contain alcohol-related data. The Directory
provides alcohol researchers with one-stop access to an index of alcohol-related
surveys and the other variables they contain.
Task 2: Data analysis and support.
This task includes preparation
of analytic reports, the Alcohol Epidemiologic Data Reference Manual, a
series of Surveillance reports, Alcohol Epidemiologic Bulletins.
Analytic reports, generally first-authored by contract staff, are submitted
to peer-reviewed journals. Recent submissions include: Alcohol and
Hepatitis C Mortality among Males and Females in the United States: A
Life Table Analysis; Residence and Alcohol-Related Health Risk
Behaviors among Undergraduate College Students; and
Alcohol Abuse
and Dependence in College and Non-College Samples: A Ten-Year Prospective
Follow-up in a National Survey.
The Alcohol Data Reference Manuals are statistical compendiums of alcohol-related
data useful to researchers and others interested in alcohol problems.
The most recent Manual, published in 2006, is Alcohol Use and Alcohol
Use Disorders in the States: Main Findings from the 2001-2002 National
Epidemiologic Survey on Alcohol and Related Conditions (NESARC). This
Manual and others including
U.S. Apparent
Consumption of Alcoholic
Beverages Based on State Sales, Taxation, or Receipt Data; U.S. Alcohol
Epidemiologic Data Reference Manual, Volume 1, Fourth Edition, June 2004;
State Trends in Drinking Behaviors, 1984-200;,
and Alcohol Consumption
and Problems in the General Population: Findings From the 1992 National
Longitudinal Alcohol Epidemiologic Survey are available on the NIAAA
website
(http://pubs.niaaa.nih.gov/publications/manual.htm).
Surveillance reports are analyses of trends based on compilations of data for
various topics. Several reports are updated annually. These include: Apparent
Per Capita Alcohol Consumption: National, State, and Regional Trends,
1977–2003; Trends in Alcohol–Related Morbidity Among Short–Stay
Community Hospital Discharges, United States, 1979–2003; Trends
in Alcohol–Related Fatal Traffic Crashes, United States, 1977–2003;
and Liver Cirrhosis Mortality in the United States, 1970–2002.
A new surveillance report Trends in Underage Drinking in the United
States, 1991-2003, will be updated every-other year. The surveillance
reports are available on the NIAAA website
(http://pubs.niaaa.nih.gov/publications/surveillance.htm).
Alcohol Epidemiologic Bulletins are short articles, written by the contractor,
for publication in NIAAA’s journal, Alcohol Research and Health.
Recent publications in Alcohol Research and Health include: Alcohol
Consumption among Young Adults ages 18 to 24 in the US: Results from NESARC;
and Gaps in Health Care Coverage for Hospitalizations with Alcohol-Related
Diagnoses.
Task 3: NIAAA-sponsored National Survey Support Services.
This task provides survey-related technical resources to the public.
In a typical 6-month period, the contractor responds to approximately
50 public requests for assistance with: access to the NESARC website,
problems with downloading data and documentation files, obtaining NESARC
publications, and technical questions. The contractor also produces CD
ROMS containing NESARC and NLAES data and documentation for public requestors.
Technical support ranges from answering questions about the survey design
and sampling characteristics of NESARC, to supplying SAS code for exceeding
low risk drinking limits.
Task 4: Technical assistance and referral services.
This task fills literature requests, provides Quick Facts for the NIAAA website,
provides technical help with poster production, and provides programming/analytic
support for NIAAA staff.
The contractor provides assistance locating alcohol-related journal articles
and books for NIAAA and the public. Using data on alcohol consumption
acquired in Task 1, the contractor produces up-to-date estimates of alcohol
consumption from nationally representative surveys including the National
Health Interview Survey and the Behavioral Risk Factor Surveillance System,
data on per capita alcohol consumption, and data on other alcohol-related
topics. These estimates are available on the NIAAA website:
http://www.niaaa.nih.gov/Resources/DatabaseResources/QuickFacts/default.htm.
In a highly collaborative relationship, the contractor assists NIAAA with the
preparation of posters, including programming and analysis as well as
graphics and printing. In the past year, the contractor assessed feasibility
for numerous study proposals by NIAAA staff (involving brief literature
searches and data exploration) and provided programming services to several
staff.
Task 5: Alcohol Epidemiologic Research Specialists
This task maintains a listing of specialized consultants and provides their services,
as needed.
In the past year, consulting services were used to review a report on
new methods for per capita alcohol consumption estimates, and to provide
statistical consultation.
Currently, DEPR is briefing all NIAAA Divisions and Research Teams about data
based in the Alcohol Epidemiologic Data Directory and reviewing with them
priority research issues to explore using the AEDS contract.
Contract cost
At issuance on 16 September 2003, estimated contract cost (exclusive of
base fee and award fee) was $5,330,770, with a base fee of $106,615 and
an award fee maximum of $426,462 for a total value over 5-years of $5,863,847.
On 3 June 2004, a formal contract modification reduced the estimated cost to
$4,809,201, with a base fee of $96,184 and a maximum award fee of $384,736
for a total value over the 5-years of $5,290,121. This represents a 10%
reduction.
Future Plans
AEDS activities will be reviewed
in 2007 by an independent advisory board. The contract expires in 2008
and will be re-competed for 2008-2013.
VII. STRATEGIC PLANNING REVIEW OF THE ALLOCATION OF DEPR’S
PORTFOLIO
For purposes of examining the DEPR portfolio, grants were coded in several ways.
Each grant was coded according to 1) the population studied; 2) whether
the study was primarily epidemiology, primarily prevention or a combination;
and 3) the problem/topic addressed. The coding for population and type
of study was mutually exclusive (i.e. only one population, and one type
of study) could be identified for each project. By contrast, the coding
for problem/topic addressed allowed for coding of projects into more than
one category since many projects address multiple problems. In addition,
for grants addressing acute and chronic alcohol-related conditions, mutually
exclusive coding for acute versus chronic condition was performed. The
results of this coding are presented in graphic or tabular form in Appendices
B-D.
POPULATIONS STUDIED
Research indicates that most alcohol-related mortality occurs among adolescents
and young adults. Further, nearly half the people ever alcohol-dependent
become dependent by age 21 and two thirds by age 25. Finally, it is clear
that the earlier youth begin to drink the greater their likelihood of
experiencing alcohol dependence, multiple dependence episodes, episodes
with more dependence symptoms as well as
alcohol-related
injuries and other problems both during their adolescent and adult years.
Therefore it is appropriate that the largest amount of the Division’s
research dollars devoted to any population group, nearly $30.6 million,
targets adolescents or college students. Another $12.4 million focuses
on women ($4.9) and minorities 9 ($7.5) while $26 million focuses on the adult population
in general. These totals are based on the mutually exclusive population
coding.
EPIDEMIOLOGY AND PREVENTION
The DEPR portfolio of research is fairly evenly balanced between epidemiology
and prevention research, an appropriate distribution.
PROBLEM/TOPIC ADDRESSED
Mutually exclusive coding revealed that the Division is currently funding 32
grants costing $12.3 million dollars that examine acute alcohol-related
conditions and 14 grants costing $6.2 million dollars that examine chronic
alcohol-related conditions. This allocation of resources is appropriate
because acute alcohol-related deaths outnumber chronic alcohol-related
deaths and produce twice the number of preventable years of life lost.
Examining the non-mutually exclusive coding of problem/topic area reveals
that grants totaling $5.8 million address motor vehicle injuries, grants
totaling $6.4 million address other unintentional injuries, grants totaling
$4.0 million address homicide and grants totaling $2.1 million address
suicide. Because motor vehicle injury deaths are the largest single category
of alcohol-related deaths, and because research in this area has translated
into substantial reductions in such deaths over time, this problem deserves
somewhat greater emphasis than it is currently receiving. Of the grants
which study chronic conditions in which alcohol is a factor, grants totaling
$5.1 million address liver disease, grants totaling $5.1 million address
cardiovascular disease, grants totaling $4.1 million address cancer, grants
totaling $3.8 million address diabetes, and grants totaling $3.9 million
address other chronic diseases. Because liver disease accounts for over
half of alcohol attributable chronic disease deaths, this area warrants
more research attention.
Among the other topic areas, grants totaling $11.3 million are studying community
interventions, grants totaling $7.8 million address public policy, grants
totaling $5.5 million address aging, grants totaling $6.0 million address
health services, grants totaling $9.8 million address measurement, grants
totaling $1.1 million address nosology, grants totaling $2.6 million address
genetic epidemiology and grants totaling $0.6 million address advertising.
Because of the recent expansion in alcohol advertising and promotion,
more attention should be focused on the effects of advertising on drinking
behavior. In addition, given the investment that has been made in the
Alcohol Policy Information System, DEPR hopes to stimulate more public
policy and health services research.
VIII. APPENDICES
A. DEPR MISSION STATEMENT
DEPR’s
Public Health Mission
- • The Division of Epidemiology and Prevention Research seeks to reduce alcohol-related
mortality and morbidity and other alcohol-related problems and consequences
through the application and integration of epidemiology and prevention
science.
Primary Goal
- • Set research priorities
- – Review and critically evaluate the state of the science
- – Evaluate and improve quality of ongoing research
- – Identify knowledge gaps and scientific opportunities
Secondary Goals
- • Stimulate and support research, training and career development by:
- – Providing technical assistance and guidance to researchers
- • In the alcohol field and other fields as appropriate
- • To NIAAA scientists
- – Working to attract new investigators to the alcohol epidemiology and
prevention fields and facilitate their progress
• Conduct research and publish in the scientific literature:
- – Secondary analyses
- – Critical literature reviews
- – Evaluation of study designs, measurement and analysis
• Promote dialogue and collaboration between DEPR and:
- – research and infrastructure teams
- – other NIAAA Divisions
- – other NIH institutes
- – other federal agencies
- – public interest groups
- – academic and research institutions
• Contribute to surveillance by:
- – Monitoring trends in alcohol consumption and problems related to consumption
- – Setting alcohol-related public health goals and tracking progress toward
meeting those goals
- – Identifying programs and policies that can help achieve those goals
- – Identifying needed but uncollected surveillance measures
- – Identifying critical research questions whose answers would further achievement
of those goals
• Disseminate information by:
- – Publishing in the scientific literature
- – Collaborating with NIAAA’s Office of Science Policy and Communications
- – Organizing scientific conferences
- – Presenting at scientific conferences
- – Participating on task forces
Two major areas of focus for the Division are:
Epidemiology of Alcohol Use and Alcohol-Related Problems
- • This broad area includes the study of:
- – Etiology (risk and protective factors) and course of alcohol-related
problems including AUDs
- – Relationship of alcohol consumption and AUDs to other diseases and disorders
- • Psychiatric comorbidity
- • Relationship to chronic diseases such as diabetes and cardiovascular disease
- • Potential health benefits of alcohol consumption
- • Relationship to sexually transmitted diseases
- – Alcohol-related consequences
- • Mortality and morbidity
- • Violence
- • Risky and unprotected sex
- • Compromised academic/vocational achievement
- • Economic costs
- – The distribution, mediation and moderation of the above by demographic
characteristics
- – Organization, financing, costs and effectiveness of alcohol-related health
and preventive services
- – Methodology
- • Nosology
- • Research design and analysis
- • Measurement (e.g. of alcohol consumption)
- • Developing and supporting innovative research approaches
The Prevention of Alcohol Use and Alcohol-Related Problems
- • This broad area includes the study of:
- – the efficacy and effectiveness of:
- • screening and brief interventions
- • prevention interventions – individual, family, school, other organization
based, comprehensive/community
- • drinking-driving countermeasures
- – the impact of:
- • public policy
- • media
- • alcohol marketing and promotion
B. GRANT AND GRANT DOLLAR ALLOCATION BY
POPULATION
