Less Drinking, Yet More Problems: Understanding African American Drinking and Related Problems

Abstract

Researchers have found that, compared to European Americans, African Americans report later initiation of drinking, lower rates of use, and lower levels of use across almost all age groups. Nevertheless, African Americans also have higher levels of alcohol problems than European Americans. After reviewing current data regarding these trends, we provide a theory to understand this apparent paradox as well as to understand variability in risk among African Americans. Certain factors appear to operate as both protective factors against heavy use and risk factors for negative consequences from use. For example, African American culture is characterized by norms against heavy alcohol use or intoxication, which protects against heavy use but which also provides within group social disapproval when use does occur. African Americans are more likely to encounter legal problems from drinking than European Americans, even at the same levels of consumption, perhaps thus resulting in reduced consumption but more problems from consumption. There appears to be one particular group of African Americans, low-income African American men, who are at the highest risk for alcoholism and related problems. We theorize that this effect is due to the complex interaction of residential discrimination, racism, age of drinking, and lack of available standard life reinforcers (e.g., stable employment and financial stability). Further empirical research will be needed to test our theories and otherwise move this important field forward. A focus on within group variation in drinking patterns and problems is necessary. We suggest several new avenues of inquiry.

Compared to European Americans, African Americans are more likely to live in poverty, be unemployed, and have lower educational attainment (U.S. Census Bureau, 2012; Williams, Mohammed, Leavell, & Collins, 2010). In the mental health field, it has often been assumed that, based on such socio-demographic factors of disadvantage, ethnic minority groups would be at greater risk for mental disorders, including substance abuse (Crum, Helzer, & Anthony, 1993; Ford et al., 2007; Gilman, Breslau, Conron, Koenen, Subramanian, & Zaslavsky, 2008; Midanik & Clark, 1995; Turner & Wallace, 2003). Interestingly, although it is true that disadvantaged social status is associated with higher risk of mental disorders, being an African American is not associated with increased risk (Kessler, Berglund, Demler, Jin, Merikangas, & Walters, 2005a; Kessler, Chiu, Demler, & Walters, 2005b). In fact, the opposite appears to be true: being African American has been found to be associated with a reduced risk of mental disorder diagnosis (Kessler at al., 2005a, 2005b; Riolo, Nguyen, Greden, & King, 2005).

Specifically pertaining to alcohol use, African Americans report abstention from alcohol at significantly higher rates than European Americans (Caetano & Clark, 1998a; Caetano & Kaskutas, 1995; Dawson, 1998; Galvan & Caetano, 2003; Substance Abuse and Mental Health Services Administration (SAMHSA), 2010, 2011). Among those who do drink, compared to European Americans, African Americans also generally report drinking less frequently and consuming smaller amounts of alcohol across almost all age groups (Caetano & Clark, 1998a; Galvan & Caetano, 2003; SAMHSA, 2010, 2011). Despite this reality, African Americans who do engage in drinking behaviors appear to be at a comparable and at times at a higher risk for experiencing alcohol related problems (Caetano, 1997; Caetano & Kaskutas, 1996; Galvan & Caetano, 2003; Jones-Webb, 1998), such as more negative social consequences from drinking (Mulia, Ye, Greenfield, & Zemore, 2009), higher rates of alcohol-related illness and injuries (Greenfield, 2001; Stinson, Dufour, Steffens, & Debakey, 1993; Yoon, Yi, Grant, & Dufour, 2001), and to some extent, alcohol dependence symptoms and/or diagnosis (Caetano, 1997; Caetano & Kaskutas, 1995; Mulia et al., 2009) compared to their European American counterparts. Although there has been an increased effort to study substance abuse in minority populations over the past two decades, no integrative theory has been developed to explain why African Americans tend to drink less than members of the dominant, European American culture, but experience higher rates of problems. In addition, most research has been conducted using a comparative research design, comparing the rate of African American use to that of the standard European American majority, thus failing to examine or explain individual differences in consumption or risk within the African American community. The aim of this paper is to offer answers to three main questions: a) Why are African Americans, in general, more likely to abstain from alcohol or drink at lower levels compared to European Americans? b) Among individuals who consume alcohol, why are African Americans more likely to experience negative consequences from drinking compared to their European American counterparts? and c) Among African Americans, who is at the greatest risk for alcoholism and alcohol related problems?

In order to answer these three questions, we have divided this paper into six sections. First, we present empirical evidence documenting lower use of alcohol among African Americans compared to European Americans. Second, we present data documenting worse alcohol-related consequences among African American drinkers compared to European American drinkers. In the third section of the paper, we provide evidence from historical, cultural, social and biological data to explain lower use of alcohol among African Americans compared to European Americans. Fourth, we then present evidence from those same perspectives to explain why there are worse consequences for use among African Americans compared to European Americans. Fifth, we present data on variations in drinking risk among African Americans, highlighting those subgroups within the African American community that may be at increased risk for problematic drinking and negative consequences from use. In the sixth and final section, we offer an integrative theory to help explain why a specific group of African Americans appear to suffer the worst consequences for drinking compared to other African Americans. We hope the theoretical framework we provide generates new lines of inquiry into this important problem and stimulates further theoretical advances.

Lower Alcohol Use among African Americans Compared to European Americans

In general, the developmental trend of alcohol consumption for Americans is a rapid rise in the frequency of drinking and the quantity consumed as one transitions through adolescence that tends to peak in early adulthood and gradually declines with increased age (e.g., Johnson, Gruenewald, Treno, & Taff, 1998). Increased consumption during adolescence is thought to be a function of experimentation with adult behaviors (Shedler & Block, 1990), peer influence to consume, and the desire to conform or fit in with one’s social network (Bray, Adams, Getz, & McQueen, 2003; Duncan, Duncan, & Strycker, 2006; Jones, Hussong, Manning, & Sterrett, 2008; Kogan, Berkel, Chen, Brody, & Murry, 2005). As individuals transition into adulthood, they experience new levels of freedom and independence (Schulenberg, O’Malley, Bachman, Wadsworth, & Johnston, 1996), which places them at increased risk for engaging in risky behaviors, such as alcohol use. As individuals move through their 20’s, the frequency and quantity of alcohol consumption tends to decrease. This change has been attributed to the increased responsibilities typically associated with adulthood, including marriage, parental, and work-related responsibilities (e.g., Donovan, Jessor, & Jessor, 1983; Jackson, Sher, Gotham, & Wood, 2001). However, much of the research presented to date has been conducted with predominantly European American samples, thus raising the question of whether this developmental trend holds for other ethnic minority groups, including African Americans.

Interestingly, developmental trends in drinking behaviors and levels of use have been shown to differ substantially between African Americans and European Americans (refer to Table 1 for a listing of studies reviewed). In comparison to their European American counterparts, African American youth report lower levels of use (Bachman et al., 1991; Johnston, O’Malley, & Bachman, 1994; O’Malley, Johnston, & Bachman, 2008; Poulin, 1991; Wallace et al., 2003a; Wallace, Brown, Bachman, & LaVeist, 2003b), begin drinking at a later age (Catalano et al., 1993; Johnson, Richter, Kleber, McLellan, & Carise, 2005; Watt, 2004), engage in less heavy drinking (Bachman et al., 1991; Wallace et al., 2003a), and show slower increases in rates of drinking in the early adolescent years (Johston et al., 1994; Warheit, Vega, Khoury, Gil, & Elfenbein, 1996) Based on a national survey that was given to youth aged 12–17, 18% of European Americans compared to 10% of African American youth reported using alcohol in the past 30 days. Moreover, 42% of European Americans compared to only 34% of African Americans reported consuming at least one alcoholic beverage in their lifetime (SAMHSA, 2010). African American youth also report higher rates of abstinence compared to European American youth (Bachman et al., 1991; SAMHSA, 2010, 2011; Wallace et al., 2003b).

Table 1.

Summary of Studies Comparing African American and European American Alcohol Consumption, by Age

Authors	Recruitment	Data Source	Sample	Gender (M/F)	Age	Results
Youth
Alcohol Initiation
Catalano et al., 1993	Community, convenience sample	Seattle, WA public schools	AA:98 EA:218	50/50	10–11	AA<EA
Johnson et al., 2005	National sample	National Drug & Alcoholism Treatment Unit Survey (NDATUS)	AA:854 EA:892	Data not provided	18+	AA<EA
Peterson et al., 1994	Convenience sample, public schools in Seattle, WA	Seattle Social Development Project (SSDP), 1988–1990	AA:142 EA:308	50/50	12–15	AA< EA
Consumption
Poulin, 1991	Community, nonrandom stratified sampling	Low income urban community, Chester, PA	AA:273 EA:111	54/46	12–25	AA<EA current
SAMHSA, 2010	National randomized sample	National Survey on Drug Use & Health (NSDUH), 2008	AA:29556 EA:169423	AA:45/55 EA:49/51	12–17	AA<EA current and lifetime
Wallace & Bachman, 1993	National, randomized sample	Monitoring the Future/ National Senior Study, 1980–1989	AA:17146 EA:119021	AA:44/56 EA:49/51	HS seniors	AA<EA current and heavy
Wallace et al., 2003	National randomized sample	MTF (1996–2000)	AA:26160 EA:140653	AA:45/56 EA:49/51	Grade 8,10,12	AA<EA current, lifetime, and heavy
Wallace & Muroff, 2002	National, randomized sample	Monitoring the Future (MTF), 1998–1999	Data not provided	Data not provided	HS seniors	AA<EA heavy
Warheit et al., 1996	Community, nonrandom sampling	Dade County, Florida Public Schools (Miami, FL)	AA:618 EA:632	79/21	10–15	AA<EA lifetime
Watt, 2004	National randomized sample	National Household Survey of Drug Abuse, 2001	AA:4557 EA:23400	50/50	12–25	AA<EA heavy
Wills et al., 2003b	Community, convenience sample	Public schools in New York metro area	AA: 343 EA: 434	Total: 53/47	Grade 7–10	AA<EA
Wu et al., 2011	National, randomized, multi-stage area probability sampling	NSDUH, 2005–2008	AA:10109 EA:43778	Data not provided	12–17	AA<EA past year
Young Adults
Consumption
Globetti et al., 1996	Community based - stratified cluster, random sampling of class sample	Mod-size public university in the South	AA:104 EA:803	49/52	college	AA<EA current
Meilman et al., 1995	Community-based randomized sample	The Core Alcohol and Drug Survey	Total:12351	Data not provided	college	AA<EA current
Peralta & Steele, 2009	Community based convenience sample	Midwestern university	AA:70 EA:259	38/62	18+	AA<EA binge AA>EA abstainer/non-binge
Siebert & Wilke, 2007	National, randomized, multistage probability sample	NHSDA, 2001	AA:931 EA:4963	AA:43/58 EA:48/52	18–21	AA<EA heavy
Adulthood
Abstinence
Caetano & Clark, 1998a	National randomized sample	Institute for Survey Research of Temple University., 1995	AA:1582 EA:1636	AA:41/59 EA:46/54	18–60	AA>EA
Caetano & Kaskutas, 1995	National, randomized multistage area probability sample	Not provided	AA:723 EA:788	AA:44/56 EA:49/51	18+	AA>EA
Lillie-Blanton et al., 1991	Community-based, multi-stage probability sampling	Baltimore Epidemiologic Catchment Area (ECA) Household Survey	AA:809 EA:1374	Females only	18–59	AA>EA
Consumption
Caetano & Clark, 1998a	National randomized sample	Institute for Survey Research of Temple University., 1995	AA:1582 EA:1636	AA:41/59 EA:46/54	18–60	AA<EA heavy
Chen & Paschall, 2003	Community convenience sample	University sample	AA: 61 EA: 391	Total: 42/58	18–25	AA<EA
Dawson, 1998	National randomized sample	National Long. Alcohol Epidemiology. Survey (NLAES), 1991	Data not provided	Data not provided	18–55	AA<EA past year
Fuchs et al., 2001	Community-based, longitudinal random sample	Atherosclerosis Risk in Communities (ARIC) Study, 1987–1995	AA:1320 EA:6817	AA:34/66 EA:45/55	45–64	AA<EA current
Gilman, 2008	Nationally representative household survey	National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), 2001–2002	AA:5826 EA:20278	Data not provided	18–60	AA<EA lifetime
Grant, 1997	National, randomized multistage sampling	NLAES, 1992	Total:42862	Data not provided	18–55	AA<EA lifetime
Herd & Grube, 1993	National randomized sample	Institute for Survey Research of Temple University, 1995	AA:635 EA:663	Females only	18+	AA<EA
Johnson et al., 2005	National sample	National Drug & Alcoholism Treatment Unit Survey (NDATUS)	AA:854 EA:892	Data not provided	18+	AA>EA later age regular use
Kandel et al., 1997	National, randomized multistage area probability sample	NHSDA, 1991–1993	AA: 20744 EA: 42206	AA:41/59 EA:46/54	12–50	AA<EA past year
Mulia et al., 2008	National randomized sample	U.S. National Alcohol Survey (NAS), 2005	AA:1054 EA:3967	AA:45/56 EA:48/52	18–50	AA<EA current
Mulia et al., 2009	National, randomized sample	National Alcohol Survey (NAS), 2005	AA:1054 EA:3967	Data not provided	18+	AA<EA current
Peralta & Steele, 2009	Community based convenience sample	Midwestern university	AA:70 EA:259	Total: 38/62	18+	AA<EA heavy
Peterson et al., 1994	Convenience sample, public schools in Seattle, WA	Seattle Social Development Project (SSDP), 1988–1990	AA:142 EA:308	Total: 50/50	18+	AA parents<EA parents current
Pletcher et al., 2005	Information not provide in article, other articles referenced	Coronary Artery Risk Development in Young Adults (CARDIA) Study, 2000–2001	AA:1374 EA:1663	AA:42/58 EA:48/52	32–47	AA<EA current AA=EA binge
SAMHSA, 2010	National randomized sample	National Survey on Drug Use & Health (NSDUH), 2009	AA:29556 EA:169423	AA:45/55 EA:49/51	12–65	AA<EA current, binge, and heavy
SAMHSA, 2011	National randomized sample	National Survey on Drug Use & Health (NSDUH), 2010	AA:30233 EA:170049	AA:45/56 EA:49/51	12–65	AA<EA current, binge, and heavy
Welte & Barnes, 1992	Community, randomized sample	New York Statewide Survey, 1986	AA: 214 EA: 87	Total: 71/29	18–34	AA>EA heavy

Note. The racial categories included are African American (AA) and European American (EA). Gender categories are male (M) and female (F). Current: current alcohol use, typically past month. Binge: binge alcohol use defined as five or more drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days. Heavy: heavy alcohol use defined as drinking five or more drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on each of 5 or more days in the past 30 days. Lifetime: defined as drinking at any time throughout life.

During young adulthood, the increase in drinking that has been shown among European American populations has also been observed among African American populations; however, rates of use and heavy drinking are significantly lower among African American young adults than among their European American peers (Caetano & Clark, 1998a; Meilman, Presley, & Cashin, 1995; Meliman, Presley, & Lyerla, 1994; Peralta & Steele, 2009; Siebert & Wilke, 2007; Strada & Donohue, 2006). Based on a national survey administered to individuals aged 18 to 25, compared to European Americans, African Americans were less likely to report being a current drinker (49.6% versus 67.7%) binge drinker (27.9% versus 46.1%), or heavy drinker (5.7% versus 17.3%; SAMHSA, 2011). Specifically related to college students, Meilman et al. (1995) and Meilman et al. (1994) found that African American college students drank less and binge drank less often than European American students regardless of whether they attended predominately European American or African American universities. Moreover, Globetti and colleagues (1996) found that among their sample of college students, approximately 60% of European American students reported that they attended parties where most or all of the time, the majority were intoxicated and nearly 36% of these students reported getting high on alcohol themselves, compared to 33% and 5%, respectively, for African American students.

Differences in developmental patterns appear to continue into adulthood. Based on a national survey administered to individuals ages 26–70, compared to European Americans, African Americans were less likely to report being a current drinker, to have consumed alcohol in the past 30 days, (46.6% versus 59.4%), a be a binge drinker (20.5% versus 22.2%), or to be a heavy drinker (4.9 % versus 6.9%; SAMHSA, 2011). In a variety of studies using different sampling methods, African American adults (Caetano, 2003; Caetano & Clark, 1998a; Caetano, Clark, & Tam, 1998; Lillie-Blanton, MacKenzi, & Anthony, 1991; SAMHSA, 2010, 2011) consistently report higher rates of abstaining from alcohol than is true for European Americans. Thus, among both nationally representative samples, as well as smaller, community based convenience samples, there is a robust overall group finding of lower rates of alcohol consumption and higher rates of abstinence among African Americans compared to European Americans across development. Moreover, these findings remained consistent, showing lower rates of drinking among African Americans, regardless of the methodology used in defining alcohol consumption (i.e., current drinker, drinking in past 30 days, lifetime, heavy alcohol use, or binge alcohol use).

Greater Negative Consequences Related to Alcohol Use among African Americans Compared to European Americans

Frequent and heavy use of alcohol has been related to several negative social, mental, and physical health outcomes, such as increased violent behavior (Rodney, Mupier, & Crafter, 1996), criminal offenses (Greenfield, 1998), legal problems (Caetano, 1997; Herd, 1997b; Jones-Webb, Hsiao, Hannan, & Caetano, 1997b), family and interpersonal problems (Caetano, 1997; Jones-Webb et al., 1997b; Mulia et al., 2009), and alcohol-related injuries and illness (Greenfield, 2001; Jones-Webb, 1998; Yoon et al., 2001). Moreover, higher average volumes of alcohol consumption are associated with increased risk for the following major chronic diseases: mouth and oropharyngeal cancer, oesophageal cancer, liver cancer, breast cancer, unipolar major depression, epilepsy, alcohol use disorders, hypertensive disease, hemorrhagic stroke, and cirrhosis of the liver (Rhem et al., 2003). Coronary heart disease (CHD) and both unintentional and intentional injuries were found to depend on patterns of drinking in addition to average volume of alcohol consumption (Rhem et al., 2003). In addition, alcohol dependence has also been linked to chronic health conditions, higher rates of alcohol-related mortality (Dawson, 2000; Rehm et al., 2003**(name)), and acute and chronic social consequences (Drummond, 1990).

In studies that exclude abstainers, there is evidence to suggest that African Americans are more likely to develop and die from liver cirrhosis and other alcohol-related causes compared to European Americans (Galvan & Caetano, 2003; Grant, 1997; Greenfield, 2001; Stinson et al. 1993; Yoon et al., 2001), despite the fact that they do not engage in heavy drinking at greater rates than European Americans (SAMHSA, 2010, 2011). For example, Buka (2002) reported that mortality from liver cirrhosis was 1.27 times more likely among African Americans than European Americans. Additionally, Kochanek and colleagues (2004) reported that mortality rates for alcohol related diseases and disorders were 10% higher in the African American population than for other ethnic groups within the United States. In line these health disparities, while it has been generally understood that there are health benefits from consuming light to moderate levels of alcohol (Byles, Young, Furuya, & Parkinson, 2006), these findings do not seem to hold among African Americans. Sempos and colleagues (2003) reported there were no beneficial effects of moderate alcohol consumption, especially for African American men. In fact, evidence appears to show that moderate consumption is associated with an increased rate of hypertension and the development of coronary calcification for African American men (Fuchs, Chambless, Welton, Nieto, & Heiss, 2001; Pletcher et al., 2005).

In line with average lower rates of use by African Americans compared to European Americans, a few studies have shown lower lifetime rates of alcohol dependence among African Americans compared to European Americans (Anthony, Warner, & Kessler, 1994; Gilman et al., 2008; Schmidt, Greenfield, & Bond, 2007). However, there have also been reports of no differences in alcohol dependence between the groups (Caetano, Baruah, & Chartier, 2011; Grant et al., 2004; Kandel, Chen, Warner, Kessler, & Grant, 1997) and most studies have found higher rates of dependence symptoms (Caetano, 1997; Caetano & Clark, 1998b; Caetano & Kaskutas, 1995; Herd, 1994a; Midanik & Clark, 1995; Mulia et al., 2009). The reason for discrepancies in comparative levels of alcohol dependence rates are unclear, and do not appear to be due to the quality of the studies, given that most are drawn from nationally representative samples. However, discrepancies may be, at least partially, due to the recentness of the data collected. Studies examining trends in prevalence of DSM alcohol dependence document that over the past decades, rates of alcohol dependence has either remained stable or declined among European Americans, but have increased among African Americans (Caetano & Clark, 1998a; Grant et al., 2004). The underlying mechanisms related to discrepancies in prevalence of alcohol dependence among African Americans and European Americans are not known and merit empirical investigation.

In addition to comparative rates of dependence symptoms and health consequences related to use, African Americans appear to have a disproportionate rate of alcohol related problems from drinking (Caetano, 1997; Grant, 1997; Herd, 1994a Jones-Webb, 1998; Mulia, Ye, Zemore, & Greenfield, 2008; Mulia et al., 2009; US Department of Health and Human Services, 1995; Wallace, 1999b). Jones-Webb and colleagues (1995) found that relative to European Americans, African Americans experienced more chronic alcohol-related problems, which were defined by the authors as both concrete drinking problems arising in various areas of the person’s life, such as financial, health, family or relatives, work, legal, etc., and alcohol dependence symptoms, as well as an increase in the percentage and magnitude of the problems across time. Mulia and colleagues (2009) also found higher rates of negative social consequences from drinking and of alcohol dependence symptoms among African Americans than among European Americans, at almost all levels of drinking, with the most pronounced differences observed at low levels of drinking. Specifically, they found that African Americans who reported low to moderate levels of heavy drinking were three times more likely to experience negative social consequences from drinking (i.e., arguments or fights, accidents, and workplace, legal, and health problems as a result of drinking) and five times more likely to report dependence symptoms (i.e., loss of control, blackouts, hands shaking, and other physiological symptoms of excessive alcohol use) compared to European Americans with similar drinking patterns.

Thus, African Americans appear to be experiencing a higher frequency of alcohol-related problems at comparable levels of use (refer to Table 2 for a listing of studies reviewed). Although the majority of studies examined within the literature review were limited in that they did not specify the specific alcohol-related problems endorsed at the highest frequency among African Americans compared to European Americans, the majority of the national randomized studies cited (8 out of 10 cited studies) reported higher endorsement of alcohol dependence symptoms among African Americans compared to European Americans. As for DSM diagnosable alcohol dependence, findings were more mixed among nationally randomized samples. There appears to be a distinction between past year dependence and lifetime, showing that African Americans appear to report lower endorsement of alcohol dependence when the criterion is lifetime dependence, but higher rates when the time frame is past year. It is unclear at this time why such differences exist; this warrants further empirical investigation. With respect to health and social problems, studies using national randomized samples consistently found higher rates of health and social problems for African Americans.

Table 2.

Summary of Studies Comparing African Americans and European Americans on Alcohol Related Problems

Authors	Recruitment	Data Source	Sample	Gender (M/F)	Age	Results
Alcohol Abuse
Grant et al., 2004	National, randomized sample	NESARC, 2001–2002; NLAES, 1991–1992	AA:756 EA:7511	AA:76/24 EA:70/30	18–65	AA<EA
Lillie-Blanton et al., 1991	Community-based, multi-stage probability sampling	Baltimore Epidemiologic Catchment Area (ECA) household survey	AA:809 EA:1374	Females only	18–59	AA>EA
Schmidt et al., 2007	National randomized sample	NAS, 2000	Data not provided	Data not provided	18+	AA<EA
Wu et al., 2011	National randomized, multi-stage area probability sampling	NSDUH, 2005–2008	AA:10109 EA:43778	Data not provided	12–17	AA<EA
Alcohol Dependence Symptoms
Caetano et al., 2001	National randomized sample	Information not provided	AA:358 EA:555	Data not provided	18+	AA>EA
Grant et al., 2004	National, randomized sample	NESARC, 2001–2002; NLAES, 1991–1992	AA:756 EA:7511	AA:76/24 EA:70/30	18–65	AA=EA
Herd, 1994	National, randomized sample	Alcohol Research Group, 1984	AA:494 EA:568	Males only	18+	AA>EA
Herd, 1997	National randomized sample	Alcohol Research Group, 1984	AA:1947 EA:1777	AA:38/62 EA:41/59	18+	AA>EA
Jones-Webb et al., 1997a	National randomized sample	National Alcohol Survey (NAS), 1992	AA:208 EA:280	Men only	18+	AA>EA
Jones-Webb et al., 1997b	National randomized sample	National Alcohol Survey (NAS), 1992	AA:436 EA:575	AA:51/49 EA:52/48	18+	AA>EA
Midanik & Clark, 1995	National randomized household probability sample	Alcohol Research Group’s 1990 survey	Data not provided	Data not provided	18–60	AA>EA
Mulia et al., 2008	National randomized sample	U.S. National Alcohol Survey (NAS), 2005	AA:1054 EA:3967	AA:45/56 EA:48/52	18–50	AA>EA
Mulia et al., 2009	National, randomized sample	National Alcohol Survey (NAS), 2005	AA:1054 EA:3967	Data not provided	18+	AA>EA
Schmidt et al., 2007	National randomized sample	NAS, 2000	Data not provided	Data not provided	18+	AA<EA
Alcohol Dependence Diagnosis
Gilman, 2008	Nationally representative household survey	National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), 2001–2002	AA:5826 EA:20278	Data not provided	18–60	AA<EA lifetime
Grant, 1997	National, randomized multistage sampling	NLAES, 1992	Total:42862	Data not provided	18–55	AA<EA lifetime AA>EA past year
Kandel et al., 1997	National, randomized multistage area probability sample	NHSDA, 1991–1993	AA:20744 EA:42206	AA:41/59 EA:46/54	12–50	AA>EA past year
Lillie-Blanton et al., 1991	Community-based, multi-stage probability sampling	Baltimore Epidemiologic Catchment Area (ECA) household survey	AA:809 EA:1374	Females only	18–59	AA>EA lifetime
Mulia et al., 2008	National randomized sample	U.S. National Alcohol Survey (NAS), 2005	AA:1054 EA:3967	AA:45/56 EA:48/52	18–50	AA>EA past year
Mulia et al., 2009	National, randomized sample	National Alcohol Survey (NAS), 2005	AA:1054 EA:3967	Data not provided	18+	AA>EA past year
Wu et al., 2011	National randomized, multi-stage area probability sampling	NSDUH, 2005–2008	AA:10109 EA:43778	Data not provided	12–17	AA<EA past year
General Problems
Caetano & Clark, 1998b	National, randomized probability sample	Information not provided	AA:1580 EA:1637	AA:41/59 EA:46/54	18+	AA>EA
Caetano & Kaskutas, 1996	National, randomized probability sample	NAS, 1992	AA:723 EA:788	AA:44/56 EA:49/51	18+	AA males >EA males
Herd, 1997	National randomized sample	Alcohol Research Group, 1984	AA:1947 EA:1777	AA:38/62 EA:41/59	18+	AA>EA
Jones-Webb et al., 1997a	National randomized sample	National Alcohol Survey (NAS), 1992	AA:208 EA:280	Men only	18+	AA>EA
Jones-Webb et al., 1997b	National randomized sample	National Alcohol Survey (NAS), 1992	AA:436 EA:575	AA:51/49 EA:52/48	18+	AA>EA
Health Problems
Fuchs et al., 2001	Community-based, longitudinal random sample	Atherosclerosis Risk in Communities (ARIC) Study, 1987–1995	AA:1320 EA:6817	AA:34/66 EA:45/55	45–64	AA>EA hypertension
Kochanek e al., 2002	National, database	U.S. Census Bureau	AA:286573 EA:1981973	AA:51/50 EA:48/52	0–85	AA>EA alcohol-induced causes of death
Pletcher et al., 2005	Information not provide in article, other articles referenced	Coronary Artery Risk Development in Young Adults (CARDIA) Study, 2000–2001	AA:1374 EA:1663	AA:42/58 EA:48/52	32–47	AA males>EA coronary calcification
Sempos et al., 2003	National randomized sample	NHANES Epidemiologic Follow-up Study (NHEFS), 1971–1992	AA: 2054	AA: 37/63	25–75	No beneficial effects of moderate drinking compared to lifetime abstainers for AA.
Stinson et al., 1993	National clinical sample	U.S. Vital Statistics published by NCHS, 1983–1989	AA:4339 EA:14689	AA:74/26 EA:76/24	15–85	AA>EA alcohol-induced causes of death
Yoon et al., 2010	National sample	National Center for Health Statistics (NCHS), 1970–2007	AA:2580 EA:22211	AA:65/35 EA:64/36	0–85	2001: AA>EA alcohol-related cirrhosis 2007: EA>AA alcohol-related cirrhosis
Social Problems
Caetano et al., 2001	National randomized sample	Information not provided	AA:358 EA:555	Data not provided	18+	AA>EA
Caetano, 1997	National, randomized probability sample	NAS, 1992	AA:723 EA:788	AA:44/56 EA:49/51	18+	AA>EA
Herd, 1994	National, randomized sample	Alcohol Research Group, 1984	AA:494 EA:568	Males only	18+	AA>EA
Herd, 1997	National randomized sample	Alcohol Research Group, 1984	AA:1947 EA:1777	AA:38/62 EA:41/59	18+	AA>EA
Jones-Webb et al., 1997a	National randomized sample	National Alcohol Survey (NAS), 1992	AA:208 EA:280	Males only	18+	AA>EA
Jones-Webb et al., 1997b	National randomized sample	National Alcohol Survey (NAS), 1992	AA:436 EA:575	AA:51/49 EA:52/48	18+	AA>EA
Midanik & Clark, 1995	National randomized household probability sample	Alcohol Research Group’s 1990 survey	Data not provided	Data not provided	18–60	AA>EA
Mulia et al., 2008	National randomized sample	U.S. National Alcohol Survey (NAS), 2005	AA:1054 EA:3967	AA:45/56 EA:48/52	18–50	AA>EA
Mulia et al., 2009	National, randomized sample	National Alcohol Survey (NAS), 2005	AA:1054 EA:3967	Data not provided	18+	AA>EA

Note. The racial categories included are African American (AA) and European American (EA). Gender categories are male (M) and female (F). Lifetime: defined as drinking at any time throughout life; Past Year: defined as consuming alcohol within the past 12 months.

Explanations for Lower Alcohol Use among African Americans Compared to European Americans

Historical Perspective

It is intriguing that in the present day, African Americans’ drinking behaviors do not appear to mirror what is found in the dominant culture. To understand the psychological and cultural factors that contribute to this phenomenon, it is important to consider the history of alcohol use among African Americans.

Records of alcohol use dating back to pre-colonial Africa indicate that the misuse of alcohol was a rare occurrence among tribal African communities. The role of beer and wine was primarily restricted to use in religious and secular ceremonies, such as rites of passage, funerals, title taking, and social gathering with the purpose of fostering a sense of community solidarity (Christmon, 1995; Harper, 1976; James & Johnson, 1996). Within each of these settings, intoxication was viewed as unacceptable behavior, with the importance of the usage of alcohol placed on participation in the ritual being conducted rather than on becoming inebriated. Among some tribes, drinking in moderation was viewed as the ideal and drunkenness was seen as a sign of weakness.

During the period of slavery in the United States, alcohol use continued to be limited among African American slaves. In part due to external pressure and control (i.e., laws prohibiting alcohol use; slave owners restricting access and use), slaves became even less likely to drink alcohol, with abstinence becoming the norm (Christmon, 1995; Herd, 1991). There have been some records of alcohol being given to slaves by slave owners; however, such use was typically restricted to holiday drinking (Christmon, 1995; Harper, 1976; James & Johnson, 1996).

The temperance movement during the 18^th century gave African Americans more reason to abstain from alcohol (Herd, 1991). Many Black leaders during the abolition and temperance movements were recorded as saying, “To keep sober was to strike a blow to slavery” (Herd, 1991). It was also recorded that prominent abolitionists encouraged abstinence among free Blacks by cautioning them to no sooner “put the intoxicating cup to his lips than he would give his back to the lash of the slave driver” (Herd, 1991). It has been noted that the growth of the African American church during this time also had a significant influence on the drinking behaviors of African Americans by promoting abstinence and moderate drinking patterns (James & Johnson, 1996). During the mid-19^th century, after the prohibition of slavery, the frequency and quantity of use among African Americans remained for the most part unchanged. According to an 1880 census report, fewer African Americans compared to European Americans died from alcoholism: 0.7 per thousand deaths for African Americans, compared to 6.7 for Irish, 2.7 for Germans, and 2.5 for other Whites (U.S. Census Office (1886) cited in Herd, 1991 and James & Johnson, 1996), providing support for the view that drinking rates were low in the African American population.

From this history, it is understandable why many African Americans abstained from alcohol or were not frequent drinkers during post-Civil War America. Importantly, the pattern of restricted use among African Americans continues to the present day. The laws prohibiting alcohol use among African Americans have been removed, and African Americans have for the most part acculturated to American culture. Thus, it is important to understand what current factors are present that place African Americans at lower risk for developing heavy episodic drinking behaviors and increase the likelihood of their abstaining from alcohol all together.

Cultural Norms and Attitudes towards Alcohol

Differences in cultural norms about alcohol use between African Americans and the dominant culture continue in present day society. Cultural norms can be understood to include both situational norms and attitudes. Situational norms have been defined as shared cultural beliefs about appropriate levels of behavior in specific contexts (e.g., Caetano & Clark, 1999; Greenfield & Room, 1997). In contrast, attitudes about drinking are general beliefs about drinking and how one personally evaluates the appropriateness of the behavior (e.g., Caetano & Clark, 1999; Fiske & Taylor, 1991). As has been true historically, for African Americans there continue to be relatively conservative norms for drinking across situations today, as well as relatively conservative attitudes toward consumption.

Among African Americans, there has been a general lack of integration of drinking into many aspects of social life, such as at regular mealtimes, religious activities, and secular occasions with political or social betterment goals (Borker, Hembrey, & Herd, 1980; Herd, 1997a; Herd & Grube, 1993, 1996). Specifically, Herd and Grube (1993) found that 80% of European American women compared to 46% of African American women reported drinking while at a restaurant at dinner. Moreover, the European American women also reported consuming over three times more than the average number of drinks consumed by African American women when at dinner at a restaurant. For adult African Americans who do drink, there also appear to be strong norms for maintaining social control and not showing signs of intoxication (Borker et al., 1980; Herd, 1997a). For example, Peralta and Steele (2009) found that compared to European American college students, African American students were less likely to drink heavily because they felt that they would be criticized for drinking more than four or more drinks in a row by their non-European American university peers. African Americans are also significantly less likely to endorse permissive attitudes about drinking, and are more likely to believe “there is nothing good about drinking” and that “drinking is not one of the pleasures of life” (Caetano & Clark, 1999; Herd, 1997a).

Parental Factors

African American youth appear to endorse similar conservative attitudes as their parents (Herd, 1994b; Ringwalt & Palmer, 1990; Wallace & Muroff, 2002). For example, they are more likely to perceive drug use as high risk and disapprove of use altogether than are European American youth (Wallace & Bachman, 1993). Three specific aspects of the African American family have been identified as protective factors against alcohol use among African American youth: providing more restricted access to alcohol, engaging in higher levels of parental monitoring of adolescent behaviors, and higher rates of sanctioning use, compared to what is observed among European American parents (James & Johnson, 1996). For example, Peterson and colleagues (1994) found that African American parents drank less (and were thus less likely to have alcohol in the home), held more negative views towards alcohol, and believed alcohol was more harmful than European American parents. Similarly, Johnson and Johnson (1999) reported that African American parents tend to be especially strict regarding alcohol use among their children, thus serving as an important buffer against use among African American adolescents. Studies have also shown that African American parents are more likely than parents of other ethnicities to closely monitor their adolescent’s activities and whereabouts (e.g., Borawski, Ievers-Landis, Lovegreen, & Trapl, 2003; Wallace & Muroff, 2002).

Racial Socialization

African American families may also be more vigilant in restricting alcohol use among their youth due to awareness of the particularly negative consequences associated with drinking for African Americans (Peterson et al., 1994). As we noted above, African Americans simply do experience more negative consequences from alcohol use (e.g., financial hardship, health problems, and problems with the law) than European Americans (Grant, 1997). The same is true for African American youth compared to their European American peers (Stewart & Power, 2002). The recognition that problems are likely to occur from consuming alcohol may motivate members of the African American community to work to restrict consumption. Thus, it may not be surprising that African American youth are not only less likely to use alcohol compared to their European American peers, but are less likely to use all types of illegal substances (i.e., marijuana, inhalant, cocaine, hallucinogen, opioid, stimulant, tranquilizer: Johnston, O’Malley, Bauchman, & Schulenberg, 1997; Wu, Woody, Yang, Pan, & Blazer, 2011).

This phenomenon may result, in part, from racial socialization. Both African American adults and youth encounter racism on a daily basis (Brody, 2006a; Kessler, Mickelson, & Williams, 1999; Landrine & Klonoff, 1996). Racial socialization refers, in part, to the process in which African American parents discuss with children the reality of negative stereotypes and racial discrimination, the value of ethnic pride, and the need to manage one’s behavior to reduce the likelihood of discriminatory actions; one example is to avoid substance use (Brody et al., 2006b; Thornton, Chatters, Taylor, & Allen, 1990).

There is evidence that racial socialization does occur among African American youth. For example, Wallace (1999a) found that African American youth were less likely to believe they could “get away with [substance use]” compared to European American youth, and perhaps accurately so. Although African Americans only represent 12% of the U.S. population, they account for approximately 50% of the juvenile correction institution population, predominately for drug possession charges and associated violence (Alexander, 1996). Conley (1994) reported that African American youth were more likely to get arrested for alcohol use compared to their European American peers, due, in part, to racial biases within law enforcement. Thus, one reason members of the African American community may constrain use is because they are attuned to the risks involved with consumption.

Religiosity

Studies have consistently shown that regardless of ethnic background, holding religious beliefs or attending religious services is associated with low rates of alcohol use and risk for problematic drinking behaviors (Brechting et al., 2010; Brown, Parks, & Zimmerman, 2001; Darrow, Russell, Cooper, Mudar, & Frone, 1992; Galen & Rogers, 2004; Wills, Gibbons, Gerrard, Murry, & Brody, 2003a). At the group level, it has been consistently shown that African Americans report higher levels of religiosity (e.g., holding religious beliefs, attending religious services, and prayer) than European Americans (Brown et al., 2001; Chatters, 2000; Chatters, Taylor, Bullard, & Jackson, 2008; Taylor, Mattis, & Chatters, 1999).

From a historical perspective, religion has consistently been an important component of daily living for individuals of African descent (Feagin & Feagin, 1999; Lincoln, 1995; Nobles, 1991). Throughout African American history in the United States, religion has played an important role as a preserver of African American heritage, an agent of reform, and an arena by which African Americans can develop and assert personal and organizational leadership skills that may be discouraged elsewhere (Billingsley, 1992; Lincoln & Mamiya, 1990; Mattis & Jagers, 2001; Pipes, 1988; Taylor et al., 1999). While millions of African Americans have embraced Christianity, with the Barna group (2009) reporting that 92% of African Americans sampled identified themselves as Christians, many other African Americans have turned to the Islamic faith. Of the approximately 7 to 8 million Muslims in the United States (Rassool, 2000), it has been estimated that 30–40% are African American (Cooper, 1999; Lumumba, 2003; Ohm, 2003).

The strong presence of religion, whether Christian, Muslim, or another faith, within the African American community may help explain mean group level differences in rates of drinking and abstinence from alcohol between African Americans and European Americans (Steinman & Zimmerman, 2004; Taylor et al., 1999; Wallace et al., 2003b; Wills, Yaegar, & Sandy, 2003b). For example, compared to European American youth, African American youth rated religion as more important, prayed more often, attended religious services more frequently, were more fundamentalist in their religious beliefs, and consumed less alcohol (Brown et al., 2001).

Biological Vulnerability and Response to Alcohol

Alcohol metabolism has been identified as one of several biological factors that can influence drinking behavior and negative health consequences from use (Yin & Agarwal, 2001). The metabolism of alcohol is primarily accomplished through two enzymes, alcohol dehydrogenase (ADH, on chromosome 4) and mitochondrial aldehyde dehydrogenase (ALDH2, on chromosome 12). ADH breaks down alcohol into acetaldehyde, which is then broken down by ALDH into acetate (Ehlers, 2007; Scott & Taylor, 2007). Both ADH and ALDH have several variations, or isoforms, which are encoded by different genes and differ in the rate at which they metabolize alcohol or acetaldehyde, respectively. Isoforms are also found at different frequencies across ethnic groups and thus may help explain variation in drinking behavior as a function of ethnicity (Osier et al., 2002).

ADH variants that metabolize alcohol more quickly and ALDH variants that metabolize acetaldehyde more slowly have been found to offer protection against alcohol dependence and heavy alcohol use (Luczak, Glatt, & Wall, 2006; Whitfield, 2002). The hypothesized mechanism for this protection is higher transient levels of acetaldehyde, which can produce stronger physiological and subjective responses at a given dose of alcohol. The most robust effects have been documented for a variant of the ALDH2 gene, ALDH2*2. Specifically, ALDH2*2 has been linked to increased sensitivity towards both the positive and negative effects of alcohol (Luczak, Elvine-Kreis, Shea, Carr, & Wall., 2002; Nishimura et al., 2001; Nishimura et al., 2002; Takeshita & Morimoto, 1999; Thomasson, Crabb, Edenberg, & Li, 1993; Wall, 2005; Wall, Thomasson, & Ehlers, 1996; Wall, Thomasson, Schuckit, & Ehlers., 1992) and lower levels of quantity and frequency of drinking (Luczak, Wall, Shea, Byun, & Carr, 2001; Takeshita & Morimoto, 1999; Wall, Shea, Chan, & Carr, 2001) among individuals of Asian descent.

Similar effects have been found for some variants of ADH genes, such as ADH1B (Cook et al., 2005; Crabb, 1995; Duranceaux et al., 2006; McCarthy, Pedersen, Lobos, Todd, & Wall, 2010). Specifically, ADH1B*3 appears to metabolize alcohol more efficiently, especially at high blood alcohol concentrations, leading to a more rapid lowering of blood alcohol levels, and simultaneously causing transiently higher levels of acetaldehyde (Ehlers et al., 2007; Lee, Hong, & Yin, 2004). There is evidence that the ADH1B*3 allele is present almost exclusively within individuals of African descent and has been estimated to be present in up to one-third of African Americans (Ehlers, Carr, Betancourt, & Montane-Jaime, 2003; Ehlers, Glider, Harris, & Carr, 2001; Luo et al., 2006; Thomasson, Beard, & Li, 1995; Wall, 2005). There is also evidence for the protective effects of the allele against alcohol dependence among African Americans based on both national (the Collaborative Study of the Genetics of Alcoholism, or COGA: Edenberg et al., 2006) and smaller community samples (n=150; Luo et al., 2006). Studies, though smaller scale convenience samples, have shown that the ADH1B*3 allele is associated with stronger positive alcohol expectancies (Ehlers et al, 2003), more pronounced physiological effects of alcohol (McCarthy et al., 2010), and negative family history of alcohol dependence (Ehlers et al., 2001). To date, no large scale genetic study has been conducted among African Americans to directly test the hypothesis that the presence of ADH1B*3 is associated with lower consumption per occasion, reduced frequency of consumption, or less frequent membership in binge drinker or heavy drinker groups. Such tests are necessary in order to understand more fully the relationship between presence of the allele and alcohol consumption for African Americans, and thus to understand factors associated with ethnic differences in alcohol consumption between African Americans and European Americans.

Summary

African culture has been characterized by cultural restraints against heavy consumption: for many hundreds of years, it has not been normative to drink heavily in this cultural group (Christmon, 1995). That history has combined with the present day experience of African Americans in the United States, characterized by high levels of scrutiny by European Americans and high levels of social sanction against boisterous behavior which together have led to lower average levels of alcohol consumption for African Americans as a group (Bonilla-Silva, 1997; Herd, 1997a). Research conducted within the past two decades, based on both national randomized and community convenience samples, has documented several culturally specific protective factors within the African American community related to reduced risk for alcoholism, such as more conservative norms and attitudes towards alcohol use, higher levels of parental monitoring of youth’s substance use behaviors, and higher levels of religiosity. These factors combine with the apparent presence of a protective genetic factor (the ADH1B*3 allele), which appears to reduce risk by accelerating the rate of metabolism of alcohol for some African Americans (Ehlers et al., 2003). Although the presence of the ADHIB*3 allele has not been researched as heavily as other alleles, such as ALDH2*2, the limited research conducted does suggest the allele offers protection against alcoholism among African Americans. However, more large scale studies need to be conducted to establish the reliability of this finding. Thus, it appears to be the case that African Americans drink less than European Americans for a combination of historical, contextual, cultural, and, perhaps, genetic reasons (Table 3 provides details on the studies reviewed in this section).

Table 3.

Summary of Studies Explaining Protective Factors of Alcohol Consumption among African Americans Compared to European Americans

Authors	Recruitment	Data Source	Sample	Gender (M/F)	Age	Results
Conservative Norms/Attitudes
Caetano & Clark, 1999	National multistage area probability sampling	Institute for Survey Research of Temple University, 1995	AA:1582 EA:1636	Data not provided	18+	AA>EA
Herd & Grube, 1993	National randomized sample	Alcohol Research Group, 1984	AA:635 EA:663	Females only	18+	AA>EA
Herd, 1994a	National, randomized sample	Alcohol Research Group, 1984	AA:494 EA:568	Males only	18+	AA>EA
Herd, 1997	National, randomized, multistage, Probability sample, 1984	Information not provided	AA:1224 EA:1034	Females only	Data not provided	AA>EA
Jones-Webb et al., 1997b	National randomized sample	National Alcohol Survey (NAS), 1992	AA:436 EA:575	AA:51/49 EA:52/48	18+	AA>EA
Peralta & Steele, 2009	Community based convenience sample	Midwestern university	AA:70 EA:259	Total: 38/62	18+	AA>EA
Peterson et al., 1994	Convenience sample, public schools in Seattle, WA	Seattle Social Development Project (SSDP), 1988–1990	AA:142 EA:308	Total: 50/50	12–15	AA>EA
Ringwalt & Palmer, 1990	Community, randomized cluster design	Public schools in North Carolina	AA:2462 EA:7407	Total: 50/50	Grade 7–12	AA>EA
Wallace & Bachman, 1993	National randomized sample	Monitoring the Future (MTF)/ National Senior Study, 1980–1989	AA:17146 EA:119021	AA:44/56 EA:49/51	Seniors	AA>EA
Wallace & Muroff, 2002	National, randomized sample	Monitoring the Future (MTF), 1998–1999	Data not provided	Data not provided	HS seniors	AA>EA
Parental Monitoring
Borawski et al., 2003	Community, convenience sample	Midwest urban high school	AA:249 EA:284	AA:23/38 EA:42/40	Grade 9–10	AA>EA negotiated unsupervised time
Religiosity
Brown et al., 2001	Community based convenience sample	Northern Ohio and Kentucky schools	AA:361 EA:500	AA:45/55 EA:47/53	14–19	AA>EA
Chatters et al., 2008	National, randomized probability sample	National Survey of American Life: Coping with Stress in the 21st Century (NSAL)	AA:3570 EA:891	AA:44/56 EA:47/53	18+	AA>EA
Taylor et al., 1999	National randomized samples	Americans Changing Lives; General Social survey; MTF; National Black Election Survey; National Survey of Black Americans	Total: 3617,26265, 16843,1151, 2107	Data not provided	Youth and adults	AA>EA
Wallace & Muroff, 2002	National, randomized sample	Monitoring the Future (MTF)	Data not provided	Data not provided	HS seniors	AA>EA
Wallace et al., 2003	National randomized sample	MTF, 1997–2001	AA:8266 EA:47738	Data not provided	Grade 10	AA>EA
Wills et al., 2003b	Community, convenience sample	Public schools in New York metro area	AA: 343 EA: 434	Total: 53/47	Grade 7–10	AA>EA religiosity AA<EA alcohol use

Note. The racial categories included are African American (AA) and European American (EA). Gender categories are male (M) and female (F). Alcohol Use: defined as consuming alcohol within the past month.

Explanations for Greater Negative Social Consequences of Alcohol Use among African Americans Compared to European Americans

Although it may seem that African Americans should be at a lowered risk for alcohol dependence and related problems given consistent findings that they engage in less heavy drinking than European Americans, the opposite appears to be true. African Americans appear to experience greater and more severe alcohol-related problems for drinking than European Americans who are drinking at comparative levels. In this next section, we summarize the body of findings that may help to explain the heightened risk of experiencing negative consequences that is found among African American drinkers.

Environmental Influences

Bonilla-Silva (1997) proposed a theory for the disproportionate level of negative consequences among African Americans. He argued that this effect is due to what he called the “racialized social system” in American society. Bonilla-Silva (1997) defines a racialized social system as a society in which the economic, political, and social structure within the society is built upon a racial hierarchy, in which some racial groups are ascribed superior positions and tend to receive better access to resources, employment opportunities, and political positions than other racial groups. One implication of a racialized social system is that the same behavior can be viewed differently as a function of one’s race. Thus, moderate drinking may not be viewed as problematic when engaged in by European American drinkers, but may be regarded as a source of concern when engaged in by African American drinkers. Thus, the higher probability of experiencing negative consequences from drinking for African Americans despite lower or equivalent use may be partly due to the racial structure of American society.

Data appear to support this view. For instance, Jones-Webb and colleagues (1997b) reported that among drinkers, although increased alcohol consumption is associated with increased negative drinking consequences (e.g., financial, health, family, work, legal problems or alcohol dependence symptoms) for European American men, there is no relationship for African American men; regardless of amount consumed, African American men are more likely to experience negative consequences for use. Moreover, Herd (1994) reported that African American men were more likely to report negative drinking consequences (e.g., dependence symptoms, problems with friends, and problems with relatives), even after accounting for drinking behavior, socioeconomic status and other social variables. It has also been observed that African Americans are more likely to be arrested for being intoxicated than European Americans regardless of the amount of alcohol consumed (Brown & Frank, 2006; Park, 1983 cited in Herd, 1994a; Neuspiel, 1996). The observation of differential treatment towards drinking based on the race of the drinker is not a new one. In 1981, Benjamin and Benjamin stated:

“moderate drinking may be tolerated if the drinker is White. A boss, foreman or policeman may ignore or simply “bear with” the White drinker, whereas the range of tolerable deviation may be narrower for the Black drinker, especially when he becomes enmeshed in the net of law enforcement. A Black drinker may become the object of racial hatred and intolerance in addition to any negative judgment toward the drinking behavior.” (p. 242)

Thus, it may be the case that studies showing higher rates of negative consequences from drinking, even at lower levels of use (e.g., Mulia et al., 2009), may be in part due to the higher likelihood of African Americans getting “in trouble” due to their skin color rather than their blood alcohol content (BAC) or behavior.

Racial discrimination

At the core of being immersed in a “racialized social system” is the influence of racial discrimination. Among all other ethnic minorities within the United States, African Americans tend to report experiencing markedly higher incidences of discrimination compared to the other groups (Landrine, Klonoff, Corral, Fernandez, & Roesch, 2006). Landrine and Klonoff (1996) found that 98% of the African American adults sampled reported that they had experienced a racist event during the past year. Racism has been conceptualized as a chronic stressor in the lives of African Americans (Clark, Anderson, Clark, & Williams, 1999), with racial discrimination being documented in almost every aspect of African American life, from home mortgage lending, to housing discrimination and residential segregation, to employment practices, to health care access and responsiveness (Smedley & Smedley, 2005).

Racial discrimination has also been identified as an important source of stress influencing health and psychological well-being (Kessler et al., 1999; Ren, Amick, & Williams, 1999; Williams & Mohammed, 2009; Williams, Neighbors, & Jackson, 2003; Williams, Yu, Jackson, & Anderson, 1997), including drinking problems among African Americans (Caetano et al., 1998; Caetano & Clark, 2000; Herd, 1994a; Gibbons et al., 2004; Gibbons et al., 2007; Gibbons et al., 2010; Jones-Webb, 1998; Jones-Webb, Hsiao, & Hannan., 1995; Kwate, Meyer, Eniola, & Dennis, 2010; Mulia et al., 2008; Taylor & Jackson, 1990; Yen, Ragland, Greiner, & Fissher, 1999). Martin and colleagues (2004) found that among employed African American men, those who reported high levels of economic distress and who perceived racial discrimination were significantly more likely to engage in stress related drinking and problem drinking behaviors. These studies suggest that although African Americans as a group drink less than European Americans, those African Americans who report more experiences of discrimination are at greater risk for problem drinking than are other African Americans.

However, not all studies have found a positive association between alcohol use and discrimination. For example, Borrell and colleagues (2007) found no relationship between drinking and self-reported experiences of discrimination among young African American adults. The authors found that African Americans were more likely to report multiple indications of extreme disadvantage than European Americans, but current drinking status and at-risk drinking were unrelated to that disadvantage. Broman (2007) found that although African American college students reported perceiving themselves as more exposed to discriminatory experiences than their European American peers, they were no more likely to engage in drinking behaviors in response to experiencing discrimination. Similarly, Kwate and colleagues (2003) found that among African American women, those who experienced less racism were more likely to be drinkers. The authors speculated that it may be the case that those individuals who have more individual experiences with racism may be more vigilant in avoiding substances that are readily connected to the negative stereotypes attributed to African Americans. Discrepancies in findings concerning the influence of racial discrimination on drinking among African Americans may also be due to the inclusion or exclusion of other covarying factors associated with drinking, such as racial identity, religiosity, and level of distress (e.g., Gibbons et al., 2010; Kwate et al., 2003; Mulia et al., 2008; Stock et al., 2011)

Residential factors/drinking context

Although African Americans, as a whole, drink less than European Americans, alcohol venues are more prevalent in African American communities (Duncan, Duncan, & Strycker, 2002; Jones-Webb et al., 2008) and lower-income neighborhoods (Bluthenthal et al., 2008; Duncan et al., 2002; LaVeist & Wallace, 2000; Romley, Cohen, Ringel, & Sturm, 2007). Furthermore, the amount of physical shelf space in convenient stores allotted for alcohol display is higher in minority and low-income neighborhoods (Bluthenthal et al., 2008). LaVeist and Wallace (2000) found that after controlling for socioeconomic status, African American residential communities had eight times as many liquor stores per capita as predominately European American communities, and the venues were often located in close proximity to African American churches, hospitals, schools, homes and recreational parks.

Thus, although African Americans are less likely to drink, for those African Americans who do consume alcohol, greater access to alcohol beverages may place them at higher risk for engaging in heavy or risky drinking behaviors. Reasons for heightened risk have been documented by several researchers. Theall and colleagues (2011) suggested that neighborhood alcohol venue density is problematic because it exposes neighborhood residents to cues related to alcohol consumption more frequently and provides residents with high levels of alcohol availability (Theall et al., 2011). Similarly, LaVeist and Wallace (2000) noted that the disproportionate concentration of liquor stores was significant because these venues typically sold larger quantities of alcohol than bars or restaurants, and the alcohol is ready for immediate consumption on street corners, nearby parks, or in motor vehicles. These concerns appear to be accurate, with evidence indicating that for African Americans who do consume alcohol, the higher concentration of liquor stores in African American neighborhoods is significantly associated with at-risk alcohol consumption (James & Johnson, 1996; Theall et al., 2011).

Additionally, drinking in low-income neighborhoods can also place African Americans at increased risk for experiencing negative consequences from drinking compared to European Americans who drink in similar contexts (Caetano & Herd, 1988 cited in Jones-Webb et al., 1995; Jones-Webb et al., 1997a). This phenomenon appears in part to be due to the cultural norm that African Americans are more likely to drink in public, such as on street corners, compared to European Americans (Herd & Grube, 1993; Nyaronga, Greenfield, & McDaniel, 2009; Stewart & Power, 2002), coupled with data indicating that police surveillance is more common in low-income African American neighborhoods than in low-income European American neighborhoods (Conley, 1994; Mastrofski, Parks, Reiss, & Worden, 1999). Thus, to the degree that African Americans consume alcohol outdoors, and to the degree that such behavior is deemed as inappropriate or excessive, drinking among this group could result in more negative consequences from drinking (e.g., citations from law enforcement) (LaVeist & Wallace, 2000; Park, 1983).

Alcohol Preference and Content

Two factors that must be considered when exploring the higher rate of negative consequences among African Americans are differences in the type of alcoholic beverage consumed and the accuracy of one’s self report of the amount of alcohol consumed. African Americans have been identified as one of the biggest consumers of malt liquor (Miller Brewing Company, 2001). Compared to lager beer, malt liquor is less expensive, has a higher alcohol content, is typically sold in larger containers (i.e., 40 oz versus 12 oz), and is packaged so that it cannot be resealed, providing a strong message to consume the beverage in one sitting (Bradizza, Collins, Vincent, & Falco, 2006; Chen & Paschall, 2003). Moreover, malt liquor is readily available and heavily marketed in low-income African American neighborhoods (Bradizza et al., 2006; Duncan et al., 2002; Herd, 2000; Jones-Webb et al., 1997a).

It is possible that some African Americans, especially youth and young adults, may be more prone to consume malt liquor over other alcoholic beverages due to the portrayal of malt liquor as a sign of masculinity among some rap artists and other African American celebrities (Chambers, 2006; Herd, 2000, 2005; Lamont & Molnar, 2001). Some researchers have suggested that this preference contributes to the higher rate of alcohol related problems in this group (Bluthenthal, Brown, Taylor, Guzman-Becerra, & Robinson, 2005; Bradizza et al., 2006; Graves & Kaskutas, 2002). However and not surprisingly, there is evidence for considerable variability in preferred alcoholic beverages among African Americans, suggesting a limited role for malt-liquor in explaining alcohol related problem rates among African Americans (Bradizza et al., 2006; Chen & Paschall, 2003; National Institute of Alcohol Abuse and Alcoholism (NIAAA), 2000; Sempos et al., 2003). To date, there have been no large population-based studies that have found a relationship between ethnic differences in alcoholic beverage preference and risk for alcohol related problems. Bradizza and colleagues (2006) studied a small sample of 53 young adults (43% African American), and found that among the entire sample, relatively few reported experiencing problems as a result of consuming malt liquor and no ethnic differences were found among those who did report problems from drinking malt liquor. There is a need for more research into this possibility before confident conclusions can be drawn.

Another possibility is that African Americans may be more likely to underestimate the amount of alcohol they consume, thus indicating a lower level of consumption than is actually the case (Kaskutas & Graves, 2000; Kerr, Patterson, & Greenfield, 2009). For example, Kerr and colleagues (2009) found when examining the type, quantity, and frequency of alcohol consumption based on ethnicity and sex, African American men were the most likely to underestimate their intake, by 31%, and had the largest overall mean drink alcohol content at 0.79 oz of alcohol, compared to 0.66 oz for the European American men sampled. The authors speculated that the higher alcohol content in African Americans’ drinks may partially explain the higher rate of alcohol problems found in this group compared to their European American counterparts. This proposed explanation also merits further empirical investigation.

Social Sanctions for Alcohol Use

A protective factor that can also work as a risk factor for negative social consequences from consumption is the conservative set of norms and values found within the African American community. As stated earlier, the African American community generally holds norms of restricted use of alcohol with discouragement of drunkenness and intoxication (Borker et al., 1980; Herd, 1997a). It follows that African Americans who do engage in heavier drinking would be subject to greater within-group negative social consequences. Indeed, Herd (1994a) found that the specific negative social consequences most commonly reported by African Americans were within-group social consequences, such as more disagreements with family and friends, trouble with one’s spouse, and disapproval of one’s drinking by peers. It may be the case that some of the alcohol dependence symptoms endorsed by African Americans are related to social disapproval of this kind.

Vulnerability to Alcohol Effects

Another protective factor that may serve as a risk factor is the genetic vulnerability to alcohol based on the presence of the ADH1B*3 allele. There is evidence that individuals with the ADH1B*3 allele who drink alcohol are more likely to report more positive expectancies for drinking, such as expectancies for enhanced sexual performance, physical/social pleasure, and increased social assertiveness (Ehlers et al., 2003). Because expectancies are understood to represent summaries of individuals’ learning histories (Bolles, 1972; Goldman, 1999; Goldman, Brown, Christiansen, & Smith., 1991), it appears that individuals with this allele may be experiencing these effects more strongly and more quickly than others, given the faster ethanol metabolism rate for these individuals. Based on this research, it may be the case that stronger expectancies experienced among African Americans with ADH1B*3 may result in more overt behaviors associated with intoxication, even at lower levels of use (Ehlers et al., 2003).

The expression of more intoxicated behaviors by some African Americans, in the face of (a) more negative within-group social consequences from intoxication and (b) more frequent surveillance of alcohol-related behaviors engaged in by African Americans on the part of law enforcement officials, may contribute to higher levels of alcohol-related problems for those African American drinkers with the ADH1B*3 allele. Thus, even though presence of the allele could protect against progression to heavier consumption, it may increase risk for problems stemming from even moderate consumption. Support for this hypothesis is provided by a study by Pedersen and McCarthy (2009), who examined the acute subjective response to alcohol among African American young adults in an alcohol administration study. They found that African American male participants experienced sharper increases in stimulation on the ascending limb of the blood alcohol curve compared to African American females at a moderate dose of alcohol. Moreover, recent work by Pedersen and McCarthy (2012) found that African Americans experienced a sharper increase in stimulation on the ascending limb compared to European Americans at the same dose of alcohol administered. These increases in stimulation were more strongly associated with higher levels of alcohol related problems, such as legal problems, blackouts, and trouble with friends for the African American participants compared to the European American participants. These findings suggest the possibility that some African Americans may have a heightened sensitivity to alcohol that result in more pronounced negative effects and negative consequences from drinking at comparable and potentially even at lower levels of use.

It is also possible that African Americans’ increased risk for mortality from liver cirrhosis (Greenfield, 2001; Stinson et al., 1993; Yoon et al., 2001) may involve the ADH1B*3 allele. Ehlers and colleagues (2007) found that in their sample of Afro-Trinidadian participants the ADH1B*3 allele served as a protective factor against alcoholism; however, among those individuals who did become alcohol dependent, the allele was associated with an enhanced risk for liver disease. This is speculated to be due to increased activity of the liver to process alcohol, which is metabolized at a faster rate among individuals with the ADH1B*3 allele (Ehlers et al., 2007).

Summary

It has been observed that African Americans report higher rates of negative consequences related to drinking compared to European Americans (studies reviewed are detailed in Table 4); this finding is particularly noteworthy because it occurs even at lower levels of use. We propose that this occurs for several reasons. First, based on a relatively small number of genetic studies using laboratory analogue designs, it appears that some African Americans may have a higher biological sensitivity to alcohol, which results in a more intense reaction to alcohol at lower quantities of use (Ehlers et al., 2003; Pedersen & McCarthy, 2012). Their stronger reaction to alcohol may lead to more alcohol-related problems, because their quicker, more intense behavior changes could result in social disapproval, both within and outside the African American community. Second, as indicated by findings from both national and community based randomized samples, cultural norms within the African American community of limited alcohol use, low tolerance for intoxication, and conservative attitudes towards drinking may result in more negative social consequences from drinking from within their own racial group than is true for European Americans (Herd, 1994a). This important finding needs to be replicated in additional samples. Third, studies that involve primarily relatively small sample, community based randomized designs have been conducted showing that African Americans are more likely to encounter legal problems and other negative consequences from drinking than European Americans, even at similar levels of use (Mulia et al., 2009). Perhaps this occurs both because of extra tight controls over African American behavior by the dominant culture, and perhaps also because the same level of alcohol consumption is likely to lead to more impairment for some African Americans than for European Americans. To date, there is little empirical information on the process by which African Americans experience more legal problems and negative consequences from similar levels of consumption; further research in this area would prove helpful.

Table 4.

Summary of Studies Explaining Risk Factors of Alcohol Problems among African Americans Compared to European Americans

Authors	Recruitment	Data Source	Sample	Gender	Age	Results
Environment
Police Presence
Conley, 1994	Community, randomized sample	Juvenile justice system	Total:41 youth	Data not provided	Data not provided	AA>EA
Mastrofski et al., 1999	Community convenience (police) and random (residents) sample	Project of Policing Neighborhoods St. Petersburg, FL	Total: 240 officers, 37 supervisors, and 1900 residents	Data not provided	Data not provided	AA>EA
Arrest for Drinking
Brown & Frank, 2006	Convenience sample	Cincinnati Police Division, 1997–1998	AA:252 EA:82	Data not provided	Data not provided	AA>EA
Caetano & Clark, 2000	National randomized sample	Information not provided	AA:1582 EA:1636	Data not provided	18–60	AA>EA
Conley, 1994	Community, randomized sample	Juvenile justice system	Total: 41 youth	Data not provided	Data not provided	AA>EA
Racial Discrimination
Borrell et al., 2007	National randomized sample	Coronary Artery Risk Dev. in Young Adults (CARDIA) Study, 1985–2001	AA:1507 EA:1813	Data not provided	18–30	AA>EA
Mulia et al., 2008	National randomized sample	U.S. National Alcohol Survey (NAS), 2005	AA:1054 EA:3967	AA:45/55 EA:48/52	18–50	AA>EA problem drinking, via racial stigma, rate decreases once adjusted for distress EA>AA problem drinking, via unfair treatment, controlling for distress
Yen et al., 1990	Community convenience sample	Muni Health and Safety Study, Transit operators; San Francisco	AA:476 EA:120	Total: 84/16	25–55	AA>EA discrimination
Residential Factors
Bluthenthal et al., 2008	Community randomized sample	Los Angeles County and Southeastern LA	AA:54 EA:62	Data not provided	N/A	↑Malt liquor shelf space=↓EA population percentage
Bradizza et al., 2006	Community convenience sample	University of Buffalo area	AA:23 EA:30	Total: 72/28	18–35	↑drinking malt liquor= ↑ enjoying outdoor activities
Duncan et al., 2002	Community, convenience sample	Metro city in Pacific Northwest	AA:486 EA:700	Total: 43/57	7–61	↑ liquor stores= ↑ poverty neighborhoods
Jones-Webb et al., 2008	National sample	Malt Liquor and Homicide Study, 2003	AA:230 EA:220	Data not provided	N/A	AA neighborhoods >EA malt liquor and exterior storefront ads for malt liquor
LaVeist & Wallace, 2000	Community, convenience sample	Baltimore, MD, 1990 US Census	AA:435720 EA:28778	Data not provided	N/A	AA neighborhoods >EA liquor stores, controlling for SES
Alcohol Content
Bluthenthal et al., 2005	Community randomized sample	South Los Angeles, Watts and surrounding areas	AA:259 EA:3	Total: 72/28	29 –50	malt liquor drinkers = ↑homeless, receive public assistance, unemployed, and drink outdoors
Bradizza et al., 2006	Community convenience sample	University of Buffalo area	AA:23 EA:30	Total: 72/28	18–35	AA=EA problems from malt liquor
Chen & Paschall, 2003	Community convenience sample	University sample	AA:61 EA:391	Total: 42/58	18–25	AA<EA drink malt liquor
Graves & Kaskutas, 2002	Community, probability sample	Clinics in Los Angeles and San Francisco Bay Area, CA	AA:185 EA:34	Females only	18+	AA>EA, females
Estimated Level of Consumption
Kerr et al., 2009	National randomized sample	NAS, 2005–2006	AA:115 EA:119	Data not provided	18+	AA>EA underestimate intake
Social Sanctioning
Herd, 1994a	National, randomized sample	Alcohol Research Group, 1984	AA:491 EA:560	Data not provided	18+	AA>EA problem with relatives and friends from drinking
Peralta & Steele, 2009	Community based convenience sample	Midwestern university	AA:70 EA:259	Total: 38/62	18+	AA>EA feel criticized for binge drinking by their non-EA peers

Note. The racial categories included are African American (AA) and European American (EA). Gender categories are male (M) and female (F). Problem drinking: defined as having experienced 1+ negative social consequences of drinking and/or multiple symptoms meeting alcohol dependence criteria as defined by the DSM-IV in the past 12 months

Up to this point, we have reviewed the literature examining group differences in drinking patterns. Of course, it is certainly true that risk levels vary within African Americans. In order to pursue a comprehensive theory of African American drinking, it is important to consider group-level processes like those considered above and also processes that contribute to individual differences in drinking behavior among African Americans. In the next section of the paper, we propose a risk model that we believe helps explain variability in drinking behavior among African Americans, in the hope that it leads to a greater focus on this aspect of African American drinking. Although this part of our model concerns individual differences, it also helps explain the apparent paradox between lower levels of drinking and higher levels of problems for African Americans.

Explaining Individual Differences in Risk among African Americans

African Americans vary in their risk for problem drinking as a function of genetic vulnerability, demographic factors, environmental factors, and learning history. Some of these risk factors (e.g., sex, family history, peer use, expectancies) have similar associations with drinking for African Americans as for European Americans. There are other risk factors that are likely to be more important for the African American community (i.e., ethnic identity, Africentrism, and exposure to discrimination) than to the European American community. Our discussion of individual differences will also include factors we have already described as important for understanding overall group differences between African Americans and European Americans; we consider them briefly again in this section because they also contribute to individual differences among African Americans (studies reviewed in this section are detailed in Table 5). We present these findings with the important caveat that, to date, there have been too few studies of individual differences within African Americans, as researchers have tended to focus on differences between races. Of the studies reviewed in this section, a majority were conducted among community-based convenience samples. Further empirical studies need to be conducted to determine the generalizability of the findings among all African Americans.

Table 5.

Summary of Studies Explaining Risk of Alcohol Consumption and Problems among African Americans

Authors	Recruitment	Data Source	Sample	Gender	Age	Results
Genetic Differences
Edenberg et al., 2006	National, randomized sample	COGA	Data not provided	Data not provided	Data not provided	ADHIB*3=↓ alcohol dependence
Ehlers et al., 2003	Community, convenience sample	San Diego county	AA: 66	32/68	18–25	ADH2*3 >ADH2*1 positive alcohol expectancies
Ehlers et al.,2001	Community, convenience sample	Information not provided	AA: 97	42/58	18–25	ADH2*3= negative family history of alcoholism.
Ehlers et al., 2007	International, community sample	West Indies	AA: 231 (East Indian or African ancestry)	88/12	18+	↑ADH1B*3 allele=↓use and alcohol dependent
Luo et al., 2006	Community convenience sample	University of CT Health Center/VA CT Healthcare System	AA: 150	Data not provided	18+	ADH1B=↓liver disease
McCarthy et al., 2010	Community convenience sample	Columbia, MO	AA: 91	42/58	21–26	ADH1B*3=↑sedation and pulse rate after alcohol consumption
Thomasson et al., 1995	Community convenience sample	5 universities in Memphis, TN	AA: 326	50/50	21+	ADH2*3>ADH2*1 faster blood ethanol disappearance rate and shorter time to reach zero blood ethanol concentration
Parental Monitoring
Herd & Grube, 1996	National, randomized sample	Alcohol Research Group, 1984	AA: 1947	Data not provided	18+	↑ involvement in black networks, ↓ heavy drinking ↑ religiosity, ↓ heavy drinking ↓pro-drinking norms, ↓heavy drinking
Clark et al., 2012	Community, convenience sample	Southeastern U.S.	AA: 567	Data not provided	9–21	↑ parental monitoring=↓30-day alcohol use
Rai et al., 2003	Community convenience Sample	Eastern urban area	AA: 1478	48/52	13–16	↑parental monitoring=↓alcohol use
Stanton et al., 2002	Community, convenience sample	Low-income areas of Baltimore City, MD	AA: 383	56/44	9–15	↑Parental monitoring=↓drug use
Tebes et al., 2011	Community, convenience sample	Northeastern U.S., afterschool program	AA: 207	54/46	6–11	↑Parental knowledge=↓substance use
Ethnic Identity
Belgrave et al., 2000	Community, convenience sample	Elementary school in urban metro area in eastern U.S., 1995–1997	AA: 195	52/ 48	8–12	↑racial identity=↓use
Brook & Pahl, 2005	Community, convenience sample	Public school in East Harlem New York City, 1990,1995,2000	AA: 333	49/51	Grade 7–10	↑ racial identity =↓ use
Burlew et al., 2000	National study	Center for Substance Abuse and Prevention (CSAP)	AA: 311	42/54	Grade 6	↑racial identity=↓ alcohol use
Caldwell et al., 2004	Community, convenience sample	Public high schools in Midwestern city, 1994–1998	AA: 488	46/54	Grade 9–12	↑racial identity=↓ alcohol use
Martin et al., 2004	National, randomized sample	National Survey of Black Workers (NSBW), 1998–2000	AA: 826	Males Only	18–96	↑racial identity= ↓alcohol use ↑Black culture=↓ alcohol use and problems
Nasim et al., 2007	Community, convenience sample	Large metro area in Northeast U.S.	AA: 114	53/47	13–20	↑racial identity=↓ heavy use
Townsend & Belgrave, 2000	Community, convenience sample	Inner-city public school, 1994–1995	AA: 104	51/49	9–12	↑racial identity=↑ conservative drug attitudes
Africentric Values
Belgrave et al., 1994	Community, convenience sample	Public school in Washington, DC	AA: 54	37/63	9.5–12.5	↑Africentric values= ↑conservative drug attitudes
Belgrave et al., 1997	Community, convenience sample	Public school in Baltimore City & Washington, DC, 1992–1994	AA: 189	44/56	8.5–13	↑Africentric values= ↑intolerant attitudes of drug use and drug harmfulness
Belgrave et al., 2000	Community, convenience sample	Elementary school in urban metro area in eastern U.S., 1995–1997	AA: 195	52/48	8–12	↑Africentric values=↑drug knowledge
Brook & Pahl, 2005	Community, convenience sample	Public school in East Harlem New York City, 1990,1995,2000	AA: 333	49/51	Grade 7–10	↑Africentric=↓drug use risk
Nasim et al., 2007	Community, convenience sample	Large metro area in Northeast U.S.	AA: 114	53/47	13–20	↑Africentric beliefs=↓alcohol use
Herd & Grube, 1996	National, randomized sample	Alcohol Research Group, 1984	AA: 1947	Data not provided	18+	↑involvement in Black networks and Black awareness=↓heavy use; via increased religiosity or decreased pro-drinking norms
Klonoff & Landrine, 1999	Community randomized sample	San Bernardino County, CA	AA: 520	47/53	18–79	↑involvement in Black networks=↓heavy use
Religiosity
Klonoff & Landrine, 1999	Community randomized sample	San Bernardino County, CA	AA: 520	243M; 277F	18–79	↑religious=↓alcohol use
Martin et al., 2004	National, randomized sample	National Survey of Black Workers (NSBW), 1998–2000	826 AA	Males only	18–96	↑church attendance and religiosity=↓consumption and problems
Nasim et al., 2007	Community, convenience sample	Large metro area in Northeast U.S.	AA: 114	53/47	13–20	↑religiosity=↓heavy use
Steinman & Zimmerman, 2004	Community, convenience sample	Public high schools in midsized city in Midwest	AA: 705	49/51	Grade 9–12	↑religious=↓alcohol use, for males
Wills et al., 2003a	Community, convenience sample	Information not provided	AA: 297	47/ 53	10–14	↑religiosity=↓substance use
Sex
Caetano & Clark, 1998a	National randomized sample	Institute for Survey Research of Temple University, 1995	AA: 1582	41/59	18–60	males>females
Ford et al., 2007	National randomized sample	National Survey of American Life (NSAL)	AA: 837	40/60	55–93	males>females
Kandel et al., 1997	National, randomized multistage area probability sample	National Survey on Drug Use and Health (NHSDA), 1991–1993	AA: 20744	AA: 41/59	12–50	males>females
Parker et al., 1995	National randomized sample	NHSDA, 1988	AA: 1888	Data not provided	12+	males>females
SAMHSA, 2010	National randomized sample	NSDUH, 2009	AA: 29556	45/55	12–65	males>females
SAMHSA, 2011	National randomized sample	NSDUH, 2010	AA: 30233	45/55	12–65	males>females
Income
Barr et al., 1993	Community randomized sample	New York State	AA: 777	43/57	18+	lower>higher males>females
Ford et al., 2007	National randomized sample	National Survey of American Life (NSAL)	AA: 837	40/60	55–93	lower>higher
Gilman et al., 2008	National randomized sample	NESARC, 2001–2002	AA: 5282	Data not provided	18–60	lower>higher
Herd, 1990	National randomized sample	Information not provided	AA: 723	Males only	18–60	lower>higher, heavy
Jones-Webb et al., 1995	National randomized sample	NAS, 1984	AA: 723	Males only	Data not provided	lower>higher, males
Age
Caetano & Clark, 1998a	National randomized sample	Institute for Survey Research of Temple University, 1995	AA: 1582	AA: 41/59	18–60	older>younger, heavy use
Herd, 1990	National randomized sample	Information not provided	AA: 723	Males only	18–60	older>younger
Johnson et al., 1998	National randomized sample	Fighting Back community evaluation, 1992–1994	AA: 4691	Total: 43/57	12–80	older>younger
Kandel et al., 1997	National, randomized multistage area probability sample	NHSDA, 1991–1993	Data not provided	Data not provided	12–50	older>younger alcohol dependence, males younger>older alcohol dependence, females

Note. The racial categories included are African American (AA) and European American (EA). Gender categories are male (M) and female (F). Current: current alcohol use, typically past month. Binge: binge alcohol use defined as five or more drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on at least 1 day in the past 30 days. Heavy: heavy alcohol use defined as drinking five or more drinks on the same occasion (i.e., at the same time or within a couple of hours of each other) on each of 5 or more days in the past 30 days. Lifetime: defined as drinking at any time throughout life.

Learning and Motives

Another set of factors that can influence risk for alcoholism are those related to psychosocial learning. One prominent psychosocial learning model emphasizes alcohol expectancies, which are thought to reflect the influence of psychosocial learning on current drinking behavior. A substantial body of research with general population samples has demonstrated that positive alcohol expectancies predict increased drinking (Barnow et al., 2004; Cumsille, Sayer, & Graham, 2000; Ouellette, Gerrard, Gibbons, & Reis-Bergan,, 1999; Settles, Cyders, & Smith, 2010; Smith, Goldman, Greenbaum, & Christiansen, 1995) and the onset of adolescent problem drinking (Christiansen, Smith, Roehling, & Goldman, 1989; Smith & Goldman, 1994). Although some studies have found no relationship between positive expectancies and drinking onset for African American youth (Chartier, Hesselbrock, & Hesselbrock, 2009), most studies have found a positive relationship that is often comparable in magnitude to what has been observed among European Americans (Flory, Lynam, Milich, Leukefeld, & Clayton, 2004; McCarthy, Miller, Smith, & Smith, 2001). It is possible that the inconsistency in findings may be in part due to differences in which additional variables are included in model tests; in particular, the inclusion of childhood conduct problems may be important. Chartier et al.’s (2009) study included the influence of both conduct disorder and positive alcohol expectancies on age of onset of drinking; they found that childhood conduct problems predicted age of onset for African American youth, but not for European American youth, whereas positive expectancies significant predicted age of onset for European American youth, but not for African American youth. Overall, it appears to be the case that at least some African American youth who have learned to associate drinking with positive consequences are more likely to drink and to drink heavily compared to African Americans who hold fewer positive expectancies about alcohol; however, this effect may depend on whether or not they experience conduct problems during childhood.

A construct related to alcohol expectancies is that of motives for drinking: most researchers understand motives to include both expected effects of alcohol consumption and one’s interest in achieving those effects (Cooper et al., 2008). Studies on adolescent motives for drinking suggest that alcohol consumption among African Americans appears to be rooted in regulation of negative mood (Bradizza, Reifman, & Barnes, 1999; Brannock, Schandler, & Oncley, 1990). Jones-Webb (1998) found that even though African American women were less likely to drink, stress had a direct and positive relationship with their drinking, especially for those who cope with stress primarily through avoidance. Thus, it appears to be the case that one influence on individual differences in drinking among African Americans is variation within that group in the motive to drink to cope with distress (Harper, 1999; Martin, Tuch, & Roman, 2003).

Environmental Risk Factors

Family factors and alcohol use

There is evidence that family composition can play an important role in alcohol initiation and use among adolescents. Donovan (2004) reported that adolescents are at a greater risk of alcohol initiation if they are living with stepparents versus intact families, if their immediate family members use substances, if they perceive parental approval or less parental disapproval of teen drinking, if they perceive lower levels of parental support, or if they perceive greater levels of parental alcohol and drug use. Specifically, Bossarte and Swahn (2008) found that adolescents who live in households with less parental supervision are more likely to engage in deviant behaviors and use substances. Likewise, greater parental monitoring is associated with decreased substance use among African American youth (Rai et al., 2003; Stanton et al., 2002; Tebes et al., 2011; Wallace & Muroff, 2002). Moreover, among adolescents who drink heavily, African American youth are significantly more likely than their European American peers to be concerned about their parents’ disapproval of their drinking alcohol (Ringwalt & Palmer, 1990). It is thought that due to the higher level of constraints placed on youth by their parents, parental influence may in fact be a stronger determinant of drug use than peer influence for African American youth compared to other racial groups (Clark, Belgrave, & Abell, 2012; Wallace & Muroff, 2002).

Another factor is exposure to violence in the home. For example, Bossarte and Swahn (2008) reported that for African American youth, exposure to violence within the home before the age of ten was associated with a threefold increase in the probability of early initiation of alcohol use. Similar findings were reported by Wallace and colleagues (1999a) for African American youth; these authors also found that the prevalence of alcohol use decreased as a function of how many parents were present in the African American adolescent’s home.

Parental substance use behaviors can also affect age of initiation and level of use among adolescents. It has been shown that having a positive family history of alcohol dependence or problem drinking increases risk for adolescent alcohol use and dependence symptoms (Chassin, Rogosch, & Barrera, 1991; Sher, 1991; Sher, Walitzer, Wood, & Brent, 1991). More broadly, children with a family history of alcohol or drug dependence are more likely to follow various deviant pathways, one of which is alcohol use (Hesselbrock & Hesselbrock, 2006; Moss, Lynch, & Hardie, 2003; Sher, Grekin, & Williams, 2005). In some studies, it has been shown that parent drinking interacts with peer behavior. For example, Jones and colleagues (2008) found no main effect for whether an adolescent consumed alcohol based on his or her parents’ level of use. However, adolescents whose peers use alcohol are at the greatest risk for drinking if their parents experienced problems from drinking and the least risk if their parents experienced few problems from drinking. This effect was observed for African Americans, as well as for European Americans.

Peer group

In general, the extent to which an adolescent’s peers engage in deviant behavior and substance use significantly influences the likelihood that the teen will also engage in such behaviors (Epstein, Botvin, Baker, & Diaz, 1999; Epstein, Williams, & Botvin, 2002; Jaccard, Blanton, & Dodge, 2005; Li, Barrera, Hops, & Fischer, 2002; Nasim et al., 2007). Specifically, peer substance use is positively associated with early age of initiation (D’Amico & McCarthy, 2006; Fite, Colder, & O’Connor, 2006; Li, Barrera, Hops, & Fischer, 2006), lifetime use (Rohde, Lewinsohn & Seeley, 1996), current use (Kuntshe & Jordan, 2006), heavy use or binge drinking (Jaccard, Blanton, & Dodge, 2005; Sher & Rutledge, 2007) and other alcohol abuse disorders (Moss et al., 2003). Bossarte and Swahn (2008) found that African American youth who reported that a few of their peers drank were 2.95 times more likely to drink than those whose peers did not drink; those who reported that most of their peers drank were 8.29 times more likely to drink. It should be noted that there is evidence that the converse is also true; perceiving less substance use among one’s peer group is associated with a lower probability of binge drinking (Stevens-Watkins & Rostosky, 2010). Specifically, the authors found that for every unit decrease in the perception of best friend’s substance use, the probability of binge drinking in high school among African American males decreased by 85%.

Ethnic identity

Ethnic identity, defined variously as self-identification with one’s ethnic group; the sense of belonging and attachment to such a group; the perceptions, behaviors, and feelings one has due to such membership; and involvement in the cultural and social practices of the group (Phinney & Kohatsu, 1997), has been associated with lower levels of substance use (Brook & Pahl, 2005; Szapocznik, Prado, Burlew, Williams, & Santisteban, 2007). Amongst African Americans, studies have consistently shown that those who strongly identify with African American culture are less likely to consume alcohol (Brook, Brook, & Pahl, 2006; Burlew, Neely, & Johnson, 2000; Caldwell, Sellers, Bernat, & Zimmerman, 2004; Herd & Grube, 1996; Klonoff & Landrine, 1999; Martin, Tuch, Roman, & Dixon, 2004; Nasim, Belgrave, Jaqgers, Wilson, & Owens, 2007) and more likely to hold conservative views toward drug use (Belgrave, Brome, & Hampton, 2000; Belgrave, Cherry, Cunningham, Letlaka-Rennert, & Phillips, 1994; Burlew, Neely, & Johnson, 2000; Klonoff & Landrine, 1999; Townsend & Belgrave, 2000). Moreover, African Americans who are more immersed in African American culture (Klonoff & Landrine, 1999), have a strong preference for African American people, and a strong connection to their families (Herd & Grube, 1996) are more likely to be abstainers.

Ethnic identity has also been identified as a buffer against substance use by mitigating the effects of racial discrimination on perceived stress, depression, problem behavior, involvement with deviant peers, and psychological well-being (Greene, Way, & Pahl, 2006; Sellers, Caldwell, Schmeelk-Cone, & Zimmerman, 2003; Stock, Gibbons, Walsh, & Gerrard, 2011; Williams, Neighbors, & Jackson, 2003). Stock and colleagues (2011) suggested that possible reasons for the buffering effect of ethnic identity against substance abuse, specifically when racial discrimination is present, include the following: a) heightened racial identity eliciting positive feelings about the self as a minority; b) endorsement of African American cultural norms, which emphasize bonds with family and racial group, which in turn, enhance feelings of belonging and social support; and c) being salient of one’s racial identity may motivate one to debunk stereotypes of African Americans as users while embracing a positive identification. Understanding the role of protective factors, such as racial identity, in the midst of risk factors, such as racial discrimination, is important in understanding the mechanisms by which certain African Americans are at increased or decreased risk for alcoholism.

Africentric world view

Holding strong Africentric world views (e.g., values, beliefs, behaviors, and consciousness deriving from an African cultural heritage) has also been associated with reduced risk of alcohol use. Nasim and colleages (2007) summarized Africentric values as reflecting the primacy of family, community, interdependent relationships and shared responsibilities. Values characteristic of an Africentric world view include: spirituality, sensitivity to emotional cues (including synthesis of the verbal and nonverbal), expressive communication (e.g., oral communication), harmony (e.g., emphasis on integrating parts of one’s life into a whole), time as a social phenomenon (e.g., fluidity of time), rhythmic movement, stylistic expression (e.g., all manners of expressing oneself), interpersonal orientation and communalism (e.g., emphasis of the group over the individual), multimodal perception and learning (e.g., perception and learning involve using visual, auditory, tactile perceptions and motor skills simultaneously, a preference for stimulus variety), and negativity to positivity (e.g., seeing the good in all situations: Randolph & Banks, 1993).

African Americans who hold stronger Africentric world views are less likely to use alcohol (Brook & Pahl, 2005) and are more likely to hold negative views towards drugs, including alcohol (Belgrave et al., 1994; Belgrave, Townsend, Cherry, & Cunningham, 1997; Belgrave et al., 2000) than are other African Americans. Likewise, Herd and Grube (1996) found that high levels of Black awareness and greater involvement in Black social networks was associated with reduced drinking and reduced heavy drinking among African Americans. When considered together, both ethnic identity and Africentric world views appear to operate as protective factors (Nasim et al., 2007): adolescents with strong ethnic identities were less likely to engage in heavy drinking even with increases in peer drinking behaviors, and holding strong Africentric beliefs was associated with later age of onset, lower lifetime use, less peer risk behaviors (i.e., having fewer peers that engaged in risky behaviors) and reduced effect of peer pressure on initiation (i.e., the more Africentric one was, the less likely peer drinking would influence alcohol initiation).

Religiosity

Holding religious beliefs appears to provide additional protection, helping to identify those African Americans who are at reduced risk for alcohol or substance use (Brook & Pahl, 2005; Nasim et al., 2007; Stevens-Watkins & Rostosky, 2010; Zimmerman & Maton, 1992). For example, Belgrave and colleagues (1997) found that among at-risk African American youth from an inner city population, those with greater spiritual beliefs (i.e., attending religious services and discussing religion or spiritual topics within the home) reported less drug use than those with fewer spiritual beliefs. Moreover, religiosity has also been shown to buffer the impact negative life events have on heavy drinking behaviors among youth in seventh through tenth grade (Wills et al., 2003b). It has also been speculated that as African Americans reach their thirties and separate more from family and parental influences and the church, their tendency to consume alcohol increases (James & Johnson, 1996).

Religious participation does also appear to be stronger among women (Barna, 2010), particularly within the African American community (Levin & Taylor, 1993; Levin, Taylor, & Chatters, 1994; Taylor, 1992; Taylor et al., 1999). The African American church may be particularly important for African American women because it provides them with a framework to reconstruct and reinterpret their roles and identities, given the multiple forms of oppression present in their lives (Grant, 1989; Levin & Taylor, 1993; McKay, 1989). Taylor and colleagues (1999) also proposed that gender differences in religiosity among African Americans may be a consequence of the gender socialization process that emphasizes many of the qualities and traits (e.g., patience, forbearance, restraint) that are consistent with religious orientation. As we discuss further below, African American women tend to drink significantly less than African American men (SAMHSA, 2009). Given that religious involvement is associated with lower rates of alcohol consumption, it is possible that sex differences in religious involvement contribute to sex differences in drinking behavior.

Demographic Variables

Sex

When comparing drinking levels among African Americans, African American women are consistently found to report lower levels of use and higher rates of abstinence compared to African American men (e.g., Caetano & Clark, 1998a; Ford et al., 2007). For example, Caetano and Clark (1998a) found that 21% of African American men compared to 10% of African American women reported drinking at least one drink in the past year. African American women also reported higher rates of abstinence (i.e., drink less than once a year or have never drunk: 55%) compared to African American men (36%). Similar findings have been reported using either convenience samples or large, representative national samples (Ford et al., 2007; Parker, Weaver, & Calhoun, 1995; SAMHSA, 2010, 2011). There is certainly heterogeneity in the drinking behavior of African American women, and researchers have begun to examine predictors and risk for alcoholism and substance abuse within this specific population (e.g., Curtis-Boles & Jenkins-Monroe, 2000). Nonetheless, African American men are, as a whole, at greater risk for problem drinking than are African American women.

Sex and income

It has been suggested that across race, income is positively associated with alcohol consumption (Grant, 1997; Moore et al., 2005; Parker, Weaver, & Calhoun, 1995; Zarkin, French, Mroz, & Bray, 1998). However, the opposite appears to be true for a subset of African American men (e.g., Jones-Webb et al., 1995; Jones-Webb et al., 1997). Studies of very poor individuals that differentiate among levels of poverty indicate that extremely poor African American men appear to be at markedly elevated risk for alcohol abuse (e.g., Ford et al., 2007; Gilman et al., 2008). Barr and colleagues (1993) found similar results: African American men who were severely impoverished (annual family income less than $15,000) were significantly more likely to report high levels of alcohol consumption than were African American women (and, interestingly, European Americans) at similar incomes. Similarly, Herd (1990) found the highest rates of heavy drinking among very low-income African American men (incomes between $10,001 and $15,000). Thus, there appears to be something specific about this subgroup of African American men that places them at higher risk of drinking, even despite their reduced income.

Not only does this subset of severely impoverished African American men appear to drink more than their counterparts, but problems from consumption appear to be more pronounced for them. African American men in lower socioeconomic classes (i.e., greater than 20% below the poverty line) were significantly more likely than comparably poor African American women (and European American men) to report greater numbers of negative drinking consequences, including problems in areas such as legal charges, finances, health, family, and work (Jones-Webb et al., 1997a). Others have reported similar findings, suggesting that African Americans who are both male and have lower income are most likely to engage in heavy or problematic drinking behaviors and experience the most problems related to use (James & Johnson, 1996; Wallace, 1999a, 1999b; Welte & Barnes, 1992). To provide a further context for our theoretical explanation of this set of findings, we next consider the joint effects of age and sex with respect to developmental trajectories of drinking.

Age, sex, and drinking trajectories

The developmental trajectory for drinking among African American men has been shown to differ significantly from their female counterparts, as well as from the drinking trajectories of male European Americans. For African American men, heavy drinking is associated with older age (Caetano, 1984; Caetano & Clark, 1998a; Caetano et al., 1998; Herd, 1990), and thus becomes more frequent than what is observed among other groups (i.e., African American women and European American men) as they age. Although somewhat dated, national survey research on heavy drinking found that for African American men, heavy drinking remained steady (although lower than European American levels) for ages 18–49, peaked during the ages 50–59, and did not drop off until the early 60s (Herd, 1990). Others have similarly noted that steady levels of drinking persist into the 50s for African American men, which is found to a much less degree for African American women (Caetano & Clark, 1998a).

To the degree that those African American men who are drinking heavily during their mid-20s continue to drink heavily well into late adulthood, they risk experiencing more severe health consequences associated with heavy drinking during this stage of life (NIAAA, 1998). Engaging in such drinking patterns may be particularly detrimental for lower income African Americans, given research indicating they tend to have less access to, and make less use of, health care, even when controlling for income and health insurance coverage compared to other ethnic groups (Schmidt et al., 2007; Weinick, Zuvekas, & Cohen, 2000).

Summary

Among African Americans, there are two subgroups that have consistently shown higher rates of heavy drinking and related problems: low-income African American men and older African American men. We next argue that the typical experience of these groups includes limited access to standard life reinforces (SLRs) and little perception that increased access to SLRs is available through sobriety. To make this argument, we will present data on the limited access to SLRs among low-income African American men compared to both low-income European American men and low-income African American women, and then present our theory as to how these factors significantly influence the risk of alcoholism and related problems among low-income and older African American men.

Theory of Heightened Risk of Drinking and Problems among Low-Income African American Men

We next present our theoretical explanation for why (a) African American men are at higher risk for problem drinking than African American women; (b) African American men of very low incomes are at the greatest risk within their ethnic group for heavy drinking and the problems that result; (c) drinking levels remain more stable later into adulthood for African American men than for other groups; and (d) low-income African American men are at the highest risk of experiencing negative health complications from drinking. The theory involves an integration of historical considerations with a novel application of a well-established, well supported theory developed from behavioral psychology.

We argue that a series of societal changes over the last century resulted in the emergence of a class of very poor African Americans, whose experiences were quite different from those of poor European Americans and wealthier African Americans. Within this group of very poor African Americans, women had continued access to a fundamentally important life role, that of mother and/or care-giver. That role enabled them to meet basic personal needs such as for relatedness and competence (Ryan & Deci, 2002) and provided them with important incentives not to pursue the reinforcement associated with heavy alcohol consumption. In contrast, very poor African American men experienced extraordinarily high rates of unemployment, fundamental disruptions in their capacity to meet the basic life needs of competence and autonomy (Ryan & Deci, 2002), very few sources of reinforcement, and hence little incentive not to pursue the reinforcement associated with heavy drinking; placing them at heightened risk for problematic drinking and related problems.

Historical Perspective

A series of social and economic events across the 20^th century led to the emergence of a class of very poor African Americans whose experiences were different from those of other African Americans and others considered disadvantaged (Farley & Frey, 1994; Massey, Gross, & Shibuya, 1994; Kasarda, 1995; Squires, Velez & Taeuber, 1991; Wilson, Tienda, & Wu, 1995). In the early 20^th century, over 1.3 million African Americans relocated from the Southern United States to the Northern, Midwestern, and Western areas to escape hardships, prejudice, and to pursue jobs, particularly those then available in factories, due to the large increase in the manufacturing industry (Christmon, 1995; Harper, 1976; James & Johnson, 1996). Policies were developed during this time to ensure the physical separation of African Americans from European Americans in residential areas, restricting housing options for African Americans to the least desirable residential areas (Cell, 1982; Fix & Struyk, 1993), thus leading to physical segregation. In the mid-1970s an era of deindustrialization took place, during which many jobs disappeared or left the inner city, resulting in a reduction in the kind of well-paying, low skill based blue-collar jobs that African American men had found when they migrated north (Squires, 1992; Wilson, 1987). As a result, many African American men who had earned good wages in the factories found that they were not well-qualified for the remaining jobs in the inner city and were often unable to afford the commute to suburban settings where many of the more plausible employment opportunities existed.

Moreover, the residential structure of inner-city communities was vastly different based on race: poor African Americans were more likely to reside in isolated poor urban neighborhoods where a majority of residents were also unemployed and impoverished, whereas poor European Americans rarely lived in such neighborhoods (Wilson, 1987). Additionally, social services were more likely to be located in predominately European American areas compared to neighborhoods with higher concentrations of African Americans (Lin & Harris, 2009). Subsequently, the inaccessibility of resources and employment opportunities, together with the spatial concentration and social isolation of poor African Americans, resulted in the emergence of an underclass African American population.

Artifacts of the residential segregation that occurred beginning in the early 1900s are still present today (e.g., Osypuk, Galea, McArdle, & Acevedo-Garcia, 2009; Williams, 1999). McKinnon and Humes (2000) reported that although Non-Hispanic European Americans account for approximately 70% of the total U.S. population, only one-fifth (21.7%) live in inner cities; in contrast, while African Americans account for about 12% of the total U.S. population, more than one-half of them (55%) live in inner cities. Acevedo-Garcia and colleagues (2008) examined opportunities for growth and development of children in the 100 largest metropolitan areas in the United States where children reside and found that 76% of African American children lived under worse conditions than the worst off European American children. Similarly, Sampson and Wilson (1995) concluded that the worst urban context in which European Americans reside is significantly better than the average context of African American communities.

In sum, African Americans have experienced dramatically different circumstances in relation to employment, adequate housing arrangements, and overall poverty than European Americans. Within this difficult context, African Americans had a reduced capacity to meet important life needs (Ryan & Deci, 2002). Furthermore, the isolation of poor African Americans from others during the deindustrialization period of the 1970s resulted in less frequent interactions with individuals or family members who resided in more stable urban areas or in the suburbs. Additionally, lower income African Americans faced more difficulties in access to welfare or social assistance programs. Together, this state of affairs describes limited access to many important life reinforcers, which can create conditions that increase the likelihood of heavy, problematic alcohol consumption. As we describe next, this lack of access to life reinforcers appears to be greater for African American men compared to their female counterparts, which places this subgroup of individuals at a heightened risk for heavy drinking and related problems.

Access to Standard Life Reinforcers

Standard life reinforcers (SLRs) have been defined as “a basic set of rewarding circumstances or experiences that persons … strive for” (Spillane & Smith, 2007 p. 405). They include housing, economic security, work opportunity, knowledge, and relationships. Spillane and Smith’s (2007) theory of how SLRs influence risk for alcoholism involves an adaptation of the behavioral choice theory (Rachlin, Logue, Gibbon, & Frankel, 1986; Vuchinich & Tucker, 1988). According to behavioral choice theory, the behavior one engages in should be understood as representing a choice among a range of possible alternative behaviors (Rachlin, Logue, Gibbon, & Frankel, 1986). One chooses a particular behavior, and the reinforcement associated with that behavior, because other reinforcers are more costly, less accessible, or unavailable. From this perspective, the choice to consume alcohol will vary as a function of (a) the constraints that are placed on access to alcohol and (b) access to and costs of alternative reinforcers (Vuchinich & Tucker, 1988). Because constraints on access to alcohol in the United States are relatively minor for adults, the availability of alternative reinforcers is thought to be central to alcohol consumption (Vuchinich & Tucker, 1988). Consistent with this view, laboratory studies have shown that alcohol consumption increases when access to alternative reinforcers is constrained, and alcohol and drug use decreases when reinforcement for drug-free activities is readily available (Carroll, 1996; Correia, Simons, Carey, & Borsari, 1998).

We believe that this behavioral choice perspective sheds light on the high level of problem drinking among low-income African American men. As a group, poor African American men have profoundly reduced access to alternative sources of reinforcement, such as financial stability (e.g., employment opportunities, equality in earnings), adequate housing, and access to social support services relative to the level of access available to European American men and African American women in similar living environments (U.S. Department of Health & Human Services, 2009; McKinnon, 2003; Stratton, 1993). Our model holds that this reality contributes to the frequency with which alcohol consumption is chosen for two reasons. First, important, alternative sources of reinforcement are less available, leading to more frequent selection of alcohol consumption as a source of reinforcement. Second, because SLRs tend not to be available, heavy drinking cannot cost access to them (e.g., there is no job to lose due to drunkenness). There is reduced disincentive to drink and hence higher levels of problem drinking. The hypothesis that limited SLR access for poor African American men contributes to their higher rates of problem drinking has not been tested, but there is clear evidence of reduced access to SLRs for this group. We next consider that evidence.

African American men versus African American women

As we noted above, African American women are less likely to consume alcohol than African American men. One reason may be increased access to some SLRs for African American women. Current research shows that among African Americans, African American women are more likely to be the sole caregiver for children (Holzer et al., 2005). McKinnon (2003) reported that 43% of African American families were maintained by women with no spouse present compared to 9% that were maintained solely by men. We believe the responsibility of caring for dependents reduces the risk of substance abuse in three ways. First, as we noted above, the role of caring for offspring enables African American women to meet the basic needs of relatedness and competence (Deci & Ryan, 2002): caring for one’s children is profoundly meaningful and hence reinforcing. Second, sobriety must be maintained to effectively care for one’s children. Third, because African American communities typically have a strong female presence, it is likely that many African American women care for other children, even if they are not their biological offspring (Burton et al., 1995; Fuller-Thomson & Minkler, 2000) and thus must maintain sobriety in order to adequately care of those who are dependent on them.

In addition, because African American women are more likely to care for dependents, they have more access to government assistance than do African American men. As limited as the government assistance is, it provides African American women a means to reinforcement that is not as readily available to African American men. Indeed, as reported by the U.S. Department of Health and Human Services, African American women compared to African American men are more likely to utilize government resources, such as the Temporary Assistance for Needy Families (TANF) program, which provides assistance in obtaining SLRs such as employment, housing, child-care, and financial stability (U.S. Department of Health & Human Services, 2009). Beginning in 1996, following welfare reform, government financial assistance became more time limited and contingent on seeking and obtaining employment (U.S. Department of Health & Human Services, 2009). For instance, TANF requirements state that “failure to participate in work requirements can result in reduction or termination of family benefits.” Participation in work requirements involves demonstrable efforts to find and maintain employment. Sobriety surely facilitates efforts to maintain employment, both in the short-term to retain TANF benefits and in the long-term once TANF assistance is finished. Indeed, empirical findings have supported the claim that TANF assistance and employment are related to reduced substance use (Corman, Dave, Richman, & Das, 2010). Thus, it appears that poor African American women face different circumstances with respect to both access to SLRs and incentives for sobriety than do poor African American men.

African American men versus European American men

Compared to European Americans, the rates of unemployment and poverty are higher among African Americans (Holzer, Offner, & Sorensen, 2005). African Americans are more likely than European Americans to be impoverished, 23% compared to 8%, respectively (McKinnon, 2003), and they experience unemployment rates that are twice as high (McKinnon, 2003; Stratton, 1993). Strikingly, over the past forty years, rates of unemployment have steadily increased among lower income African American men (Western & Pettit, 2000). Consistent with these findings, a Wall Street Journal analysis of over 35,000 U.S. companies found that African Americans were the only racial group to experience a net job loss during the 1990–1991 economic downturn: African Americans had a net job loss of 59,000 jobs, compared with a net gain of 71,000 for European Americans, 55,100 for Asian Americans, and 60,000 for Latinos (Sharpe, 1993 cited in Williams, 1999).

European American men have a better chance of employment, even at similarly low income and educational levels as African American men. Considering 20-year old high school dropouts living in low-income areas, unemployment rates were found to be 62.8% for African American men compared to 24.5% for European American men (Stratton, 1993). In experimental research in which the name on a job resume is varied to “sound African American” or “sound European American” (e.g., Jamal versus Greg; Lakisha versus Emily), European American sounding names (using identical resumes) yielded as many more callbacks as an additional eight years of work experience. Moreover, for those men who do find employment, European American men are likely to have higher earnings than African American men at similar positions (Cohen, 1998; Eller, 1994; Orzechowski, 2003).

Another issue concerning access to employment is that of incarceration. It has been well documented that African American men are overrepresented within the penal system (e.g., Bureau of Justice Statistics, 2003). Western and Pettit (2000) reported that in the mid-1990s, just under half of all male convicts were African American men. The probability of being convicted of a felony is higher for African Americans than European Americans: African Americans comprise 12% (versus 82% for European Americans) of the adult population in the United States, but make up 38% (versus 49% for European Americans) of those convicted of felonies within the prison system (Rosenmerkel, Durose, & Farole, 2009).

Among individuals with a criminal record, African Americans have more difficulty obtaining employment than do European Americans (Holzer et al., 2005). Pager (2003) found striking racial differences in the rate of callbacks among both convicted and non-convicted individuals: African Americans without a criminal record received callbacks 14% of the time, compared to 34% for European Americans. Among those with criminal records, rates of callbacks were 5% for African Americans compared to 17% for European Americans. Strikingly, European Americans with criminal records received more favorable responses (17%) than African Americans with no criminal record (14%).

SLR access, race, sex, and age

As documented above, access to the SLR of employment appears to be greater for low-income African American women and European American men than for low-income African American males. It is also plausible that because the limited access to SLRs is not likely to improve with age, this subgroup of African American men are also more likely to drink and drink heavily well into later adulthood (James & Johnson, 1996). This may partially account for the stable percentage of heavy alcohol drinkers found among African American men into later adulthood (SAMSHA, 2010). These percentages of heavy alcohol consumption outnumber both African American women and European American men and women, especially in later adulthood (SAMSHA, 2010). Although no longitudinal study examining the course of alcohol use among African American men with extreme disadvantage (i.e., limited SLR access) from adolescence to later adulthood has been conducted, findings from cross-sectional studies provide a partial picture of the risk that is consistent with our model. Lifetime substance use among African Americans is significantly associated with increasing age, male sex, little education (i.e., having less than 12 years of education), and low income (Ford et al., 2007; Gillman et al., 2008; Herd, 1990).

Problems associated with the higher rates of substance use in this group appear to be compounded by reduced access to and/or reduced utilization of health care services within low-income neighborhoods (e.g., Lin & Harris, 2009). Poor African American men drink more, they have few disincentives to drink, they drink while older, and they tend not to utilize health care services. This subgroup of African Americans appears to be at increased risk for enduring prolonged and potentially fatal consequences from their alcohol use. Alcohol-related liver disease mortality rates have been found to be considerably higher among African American men (7.4 per 100,000) compared to European American men (5.2 per 100,000: Kim, Brown, Terrault, & El-Serag, 2003; Stinson, Nephew, Dufour, & Grant, 1996), and the highest level of risk was for those African American men who were single, urban residents, unemployed, and had low educational attainment and family income (Singh & Hoyert, 2000). It is possible that increased access to, and utilization of, health care could mitigate these effects.

Summary

Although there is certainly a need for more national, randomized sample studies of this problem and longitudinal studies directly examining the complex interplay of access to SLRs on drinking behaviors among African American men across development, it appears to be the case that very impoverished and older African American men represent one group of African Americans at high risk for problem drinking (Jones-Webb et al., 1997a; Wallace, 1999b). We theorize that lower-income African American men are at increased risk due to (a) the experiences of distress associated with racial inequities; (b) lack of accessibility to important life reinforcers, such as employment, financial stability, adequate housing, and responsibility for child-care; and thus (c) difficulty meeting basic needs for competence and autonomy. We suggest that because these factors are likely to persist over time, there is a higher rate of heavy drinking in older low-income African American men, thus placing them at further increased risk for health problems from their drinking (NIAAA, 1998).

CONCLUSIONS AND IMPLICATIONS

Conclusions

Compared to their European American counterparts, African Americans, when considered as an overall group, report higher rates of abstinence and report lower rates of alcohol use based on both current and lifetime use (SAMHSA, 2010, 2011). These findings hold across development, with African Americans reporting lower rates of use and later initiation of drinking during adolescence, report lower rates of drinking and binge drinking during college, and report lower rates of heavy drinking into adulthood (Caetano & Clark, 1998a; Faden, 2006; Wallace et al., 2003a). Despite lower rates of this risky behavior, African Americans, as a whole, report higher rates of alcohol related problems (Mulia et al., 2009; Yoon et al., 2001). Specifically, they report more negative social consequences from drinking, higher rates of alcohol-related illness and injuries, and, possibly, higher rates of alcohol dependence diagnosis and symptoms compared to European Americans.

Based on an extensive review of literature in the areas of history, sociology, behavioral psychology, and biology, we help to explain the following: a) Why African Americans are more likely to abstain from alcohol or drink at lower levels compared to European Americans; b) Among individuals who drink, why African American drinkers are more likely to experience negative consequences from drinking compared to European Americans; and c) Which African American subgroup is at the greatest risk for alcoholism and alcohol related problems. In short, African Americans have a long history of abstinence or restricted use of alcohol dating back to African culture, pre-slavery. These practices have been maintained in the present day due to a combination of cultural norms and attitudes toward restricted use, high regard to religious beliefs and customs disapproving of alcohol use, and genetic vulnerability limiting the amount of alcohol necessary to experience the effects of alcohol. Evidence indicates that the negative consequences experienced by African Americans who do consume alcohol are at least in part due to (a) stronger concerns about African American drinking than European American drinking by the dominant society and law enforcement; (b) a higher level of social sanctions for use within the African American community, which, though a protective factor, appears to result in more within-group negative social consequences for African American drinkers; (c) characteristics of typical drinking contexts and the location of liquor stores; (d) possible underestimates of amount consumed by African Americans; and (e) a genetic vulnerability that, though protective, may result in signs of intoxication after only moderate consumption, thus triggering both within-group and societal social sanctions.

To address the third question, which African American subgroup is at the greatest risk for alcoholism and alcohol related problems, we began by examining those risk factors that help explain variability in alcohol use within the African American community. We felt that doing so was crucial; the past emphasis on contrasting African Americans and European Americans provides little information on who is at particular risk within the African American community. We found clear evidence that very low income African American men appeared to be at the highest risk for heavy problematic drinking and, in turn, for experiencing the highest rate of negative problems from drinking. We invoked behavioral choice theory (Rachlin et al., 1986; Vuchinich & Tucker, 1988) to help explain these findings. We theorized that, because these men have restricted or limited access to standard life reinforcers (SLRs), such as adequate employment opportunities, financial stability, adequate housing, and responsibility for health care, they (a) are more likely to pursue the reinforcement associated with heavy drinking and (b) have fewer disincentives for such heavy drinking. We theorized that the high level of negative health consequences found among African Americans may be due to two factors. The first is that the heavy drinking characteristic of some members of this group is sustained into late middle age. If findings suggesting an association between the ADH1B*3 allele and negative health complications are accurate, then the subset of these individuals who have the genetic vulnerability to alcohol are a significantly high risk of liver disease and cirrhosis (Ehlers et al., 2007). The second is that, overall, African Americans are less likely to have health care or utilize health services, such that if problems were to develop from use, they would be less likely to get needed treatment, resulting in higher severity and frequency of health problems from use (Schmidt et al., 2007). Overall, there appear to be several social, biological, and behavioral factors that place African Americans at higher risk of more severe consequences from drinking than European Americans, and for some the consequences are present regardless of level of consumption.

Implications

We believe there are a number of implications of the theory we have offered, and we hope this paper proves generative of theoretical advances beyond what we have offered as well as new, informative empirical research. Many further theoretical advances are necessary to improve our understanding of African American drinking. For example, there is considerable room to elaborate on the process by which individuals choose the reinforcement associated with alcohol use when faced with limited access to SLRs. One body of theory and research that may prove relevant is that concerning the pursuit of immediate versus delayed rewards. One important skill for adaptive functioning is to have the capacity not to respond to immediate urges or immediately available reinforcement, if doing so undermines one’s important long-term goals and interests (Davidson, 2003). Developmental science emphasizes the ongoing formation of the capacity to organize one’s behavior around important goals (Lerner, Freund, Stefanis, & Habermas, 2001). When, as we have argued here, circumstances are such that one does not perceive a way to pursue important, long-term interests (one sees limited access to SLRs), perhaps what is thought of as a normal developmental process of learning to pursue long-term goals does not occur in the same way. In the absence of perceived access to both immediate and long-term rewards, it is perhaps unlikely that one develops the skills necessary to balance pursuit of the two types of rewards. If so, then a crucially important aspect of prevention might involve provision of access to both forms of reward. Whether this can occur on an individual level, or instead requires societal change, is unclear.

A number of implications of the theory we have offered can be tested empirically. Tests of the theory would require epidemiological research, laboratory studies, and field studies. For example, one implication of our model is that changes in unemployment among African American males should alter perceived access to SLRs, and thus alter drinking levels. The U.S. experienced a sharp increase in the overall unemployment rate in the late 2000’s, and the unemployment rate was considerably higher for African American males (U.S. Bureau of Labor Statistics, 2011). Our theory would suggest that one result of this trend would be a sharper increase in alcohol consumption among unemployed African American males compared to unemployed European American males and African American females.

Moreover, based on theory, we would speculate that a decrease in access to SLRs among any subgroup of individuals, regardless of ethnicity, would also be associated with increased risk for alcoholism (Caetano & Clark, 1998a). This has been shown among substance using women. For example, Turner and Wallace (2003) found that compared to women who do not report substance use, drug-using women are more likely to be single, separated, or divorced; have less than a high school education; and have fewer sources of social support. Moreover, Turner and Wallace (2003) speculated that the relationship between stressful life events and risk behaviors may be intensified in low-income, ethnic minority groups. We would hypothesize that this is due to more limited access to SLRs, as well as the compounded effect of racism and poverty.

Laboratory studies are needed to further investigate response to alcohol in African Americans. For example, we reviewed evidence that African Americans experience a sharper increase in stimulation compared to European Americans. In European Americans, experiencing increased response to the rewarding effects of alcohol (e.g., stimulation) is related to experiencing increased behavioral disinhibition while intoxicated (e.g., Assaad et al., 2006). Further, increased behavioral disinhibition while intoxicated has been related to increased alcohol use (Marczinski, Combs, & Fillmore, 2007). African Americans may experience more alcohol problems at lower blood alcohol concentrations as a result of increased response to alcohol leading to reduced behavioral control. This important pathway has not been examined in African Americans.

Research on pharmacological interventions for problem drinking among African Americans may be important. To the degree that some African Americans experience more stimulation from alcohol, pharmacological agents that reduce alcohol’s reinforcing or pleasurable properties might prove beneficial for those individuals. Interestingly, Naltrexone, which is thought to have those effects, has not reduced stimulation levels among African American drinkers (Plebani, Oslin, & Lynch, 2011; Ray & Olsin, 2009); there is a clear need for further research to develop effective interventions for this group.

Contextual factors such as where African Americans drink, what type of alcohol they drink and their experiences while they are drinking need to be further examined, in part with respect to how they influence African Americans’ responses to alcohol. This research would be informative both through laboratory and field studies. The field studies could be particularly useful to test implications of the heightened responsivity to alcohol experienced by some African Americans: perhaps heightened responsivity implies faster learning about alcohol’s effects, such that during the earliest exposure to alcohol African American children will form stronger expectancy-drinking relationships than European American children. If that does occur, is it the case that conservative African American cultural norms work against such early expectancy formation to mitigate risk?

In relation to problems experienced by African Americans, based on our theory, we speculate that at least some of the problems experienced by African Americans are more social in nature (e.g., disapproval from peer and family for drinking alcohol and/or becoming intoxicated), and may be less likely due to very heavy levels of consumption or physiological problems (i.e., increased tolerance and withdrawal symptoms) from alcohol use. For example, a study by Keyes and colleagues (2009) found that when weekly at-risk drinking was included as a necessary condition for an alcohol dependence diagnosis, risk for dependence was significantly reduced for African Americans, controlling for socioeconomic status. Moreover, a study by Caetano, Clark, and Greenfield (1998) reported that among treatment seeking individuals, African American men reported the lowest prevalence of withdrawal symptoms, preceded by European American and Hispanic men. Thus, there appears to be some evidence that the problems experienced by African American drinkers may be less physiological and more social in nature. There is a need for further research to examine the nature of the alcohol dependence symptoms endorsed by African Americans versus those endorsed by other groups. If a difference does exist, then different treatment interventions may be necessary in order to address the specific problems that are experienced by African Americans who do engage in drinking behaviors.

Can differences in cultural norms be measured as group differences in motives not to drink (Metrik, McCarthy, Frissell, MacPherson, & Brown, 2004)? Does the nature of motives not to drink differ between the two ethnic groups (Cooper et al., 2008)? Are there longer-term goals that are valued by African Americans that provide motives not to engage in heavy consumption, such as a motive to be faithful to one’s religious commitments? Are there other such goals? If cultural norms regarding alcohol among African Americans, perhaps expressed as motives not to drink, are more healthy than European American norms, can they be used to improve prevention programs for other groups? How might that occur? How does the current analysis inform efforts to understand alcohol-related behavior in other ethnic groups?

Advances in many relevant fields have made it possible to develop this integrative theory of risk. We believe researchers are closer than ever to understanding the complexity of African American drinking, and we offer this theory in order to stimulate new research and further theoretical advances so that this important problem can be understood.