The prevalence estimates and adjusted ORs for current, heavy, and binge drinking were higher among Hispanics who chose to complete the BRFSS in English than those who chose to complete the survey in Spanish. These results held true for both men and women separately, although the differences were particularly striking among women. To our knowledge, these are the most recent, nationally representative data stratified by language of survey on drinking patterns among Hispanics in the United States, and the results suggest that preferred language of survey administration is predictive of alcohol use in this population.
The finding that English-speaking Hispanic women were more likely to drink than were English-speaking Hispanic men confirms previous research on acculturation and substance abuse (17
). Alcohol use has been a part of American culture for more than 300 years, and its use is a socially accepted behavior (33
). As Hispanic women become more acculturated to American society, they are possibly more willing to participate in the social norms of the host society and less likely to feel the influence of traditional Hispanic culture (11
). As has been previously noted, acculturated individuals tend to have more liberal attitudes (23
). Therefore, they are more willing to participate in behaviors that previously had been taboo for them. Alternatively, Hispanic men in their native culture may have felt less societal pressure to abstain from drinking (36
). These ideas could possibly explain why women who chose to complete the survey in English demonstrated a stronger proclivity to participate in drinking than did men.
Variations in drinking patterns among the heterogeneous Hispanic populations in the United States have been attributed to sex (37
), country of origin (6
), and level of acculturation (38
). Acculturation has been a prominent focus of research on alcohol use among Hispanics (40
). Acculturation involves changes in the beliefs, attitudes, and behaviors of immigrant populations as they adapt or assimilate to living in the dominant culture or society. Measurements of acculturation have typically included several questions such as country of origin, length of time spent in the host country, language preference, and feelings of interaction with the new culture (42
Measuring acculturation in large population-based surveys such as the BRFSS can be challenging because the instruments used to measure the construct are lengthy. Large national surveys such as the BRFSS and the National Health Interview Survey, for example, collect data on many aspects of health and health care, which means there is competition for question space within the surveys. Given this challenge, the use of a 1-item question on language preference as a proxy measure may provide a feasible and accurate method to assess acculturation. A previous study of cigarette smoking behavior among US Latino men and women reported a high correlation (r
= 0.8) between the 1-item question on language preference for the survey and a validated instrument used to measure acculturation (44
Furthermore, researchers using the BRFSS in Oregon tested the 1-item question on language of survey to examine the effect of acculturation on smoking (20
). Given that this measure has been used previously as a predictor of acculturation and was found to provide similar results to studies using more in-depth measures of acculturation, the use of this 1-item proxy in a national survey could save both time and money when conducting large numbers of interviews.
However, we are not advocating an indiscriminate use of proxy measures of acculturation in all research. Much research focusing specifically on the effects of acculturation and drinking should continue to employ longer and multi-item measures of acculturation. This will allow for a deeper understanding of how different dimensions of this construct (eg, language use, adoption of social norms, social interaction patterns, access to employment) are associated with drinking behaviors and how such associations change with different alcohol-related behaviors cross-sectionally and over time.
One practical implication of our study is the need to create materials in Spanish to reduce or prevent alcohol misuse. Alcohol industry representatives, noting the increasing number of Hispanics in the United States, have targeted this population (45
). Alcohol product advertising and marketing campaigns in Spanish have been implemented, and a substantial proportion of Hispanic youth is being exposed to both English and Spanish alcohol advertisements (45
). Therefore, from a public health perspective, Spanish language materials that discuss the effects of alcohol use should be used to target both Hispanic adults and youth. Providing interventions in Spanish may help the intervention more fully resonate with the Hispanic population and possibly provide primary prevention against alcohol misuse among Hispanics early in the acculturation process.
In our review of the literature, we found no studies that addressed preventing Hispanic women from adopting US alcohol use patterns during the acculturation process. Our study and others (42
) suggest that interventions and further research in this area are needed.
Our study had some limitations that should be considered. First, the BRFSS is a land-line telephone survey, and not all people in the United States possess land-line telephones. A report from the US Census Bureau indicated that Hispanic households have a slightly lower rate of land-line telephone coverage than do white households (46
). This situation could reduce the likelihood of Hispanics participating in this survey. Second, because findings are based on self-report, response bias may have been introduced into the results. Furthermore, underreporting of alcohol consumption is especially common among survey participants (47
). Therefore, these estimates of drinking behaviors may be lower than the actual occurrence.
Despite these limitations, our analyses suggest that language of preference for survey interviews is associated with alcohol use among Hispanics in the United States. As previous work has suggested, language preference alone may be used as a reasonable proxy measure of acculturation. Future studies should consider language preference and its association with other health behaviors among Hispanics in the United States.