The American Journal on Addictions, 20: 420–428, 2011 C American Academy of Addiction Psychiatry Copyright ISSN: 1055-0496 print / 1521-0391 online DOI: 10.1111/j.1521-0391.2011.00149.x
Psychological Distress, Binge/Heavy Drinking, and Gender Differences among Older Adults Namkee G. Choi, PhD, Diana M. DiNitto, PhD School of Social Work, University of Texas at Austin, Austin, Texas
Previous studies provide conflicting information on the relationship between drinking and mental health problems in older men and women. This paper addresses the relationship of binge/heavy drinking to psychological distress in community-dwelling older men and women. The study sample consisted of subjects aged 50 and older (n = 2,462 men and 2,863 women) who participated in the 2008 National Survey on Drug Use and Health. Men and women who drank and drank heavily were compared, and ordinary least squares regression analysis was used to examine the association between drinking and psychological distress after demographic factors and health status were controlled for. Binge/heavy drinking was found to be related to psychological distress in older women, but not in older men. Though findings on the relationship between heavy drinking and psychological distress in men and women are not consistent across studies, older men and women should be assessed for both conditions in clinical settings. (Am J Addict 2011;20:420–428)
INTRODUCTION Health and social service providers working with older adults should be concerned about drinking, psychological distress, and the relationship between the two. Though drinking tends to decrease with age,1,2 significant numbers of older adults, particularly men, exceed recommended drinking limits. For example, in a sample of more than 12,000 Medicare beneficiaries, 16% of men and 4% of women were found to engage in an unhealthy level of drinking.3 Rates of mental health problems, on the other hand, have been found to be higher among older women. For example, the National Comorbidity Survey Replication (NCS-R) found that while the likelihood of having a mood, anxiety, or combined mood–anxiety disorder showed a pattern of decline with age, the 12-month rates of disorders Received June 7, 2010; revised August 10, 2010; accepted September 2, 2010. Address correspondence to Dr. Choi, School of Social Work, University of Texas at Austin, 1 University Station, D3500, Austin, TX 78712–0358. E-mail: [email protected]
were significantly higher among women aged 55 and older than men in the same age group.4 Another study using NCS-R data also showed that women in each age group, including 65 and older group, were more likely than their male counterparts to have had a major depressive episode.5 A study based on the National Survey of American Life also found that African American women aged 55 and older had a higher rate of lifetime anxiety disorder than their male counterparts, though the men had a higher rate of lifetime substance use disorders and no gender difference was found in any 12-month disorder that included substance use disorder.6 Given the association between at-risk or problem drinking with higher mortality and physical and mental problems among older adults,7–9 and considering the differences in rates of at-risk or problem drinking and mental health problems generally found among older men and women, we sought to explore these relationships further. We were also prompted to conduct this study because previous studies have produced conflicting results on the relationship between problem drinking and depression/other psychiatric disorders among older men and women. For example, cross-sectional studies have found a relationship between alcohol misuse and depression in both men and women.10–12 However, in a study of lifetime mental disorders based on Epidemiological Catchment Area data, depressive symptoms were more strongly associated with the development of alcohol dependence in women than in men; in addition, alcohol symptom severity was more strongly associated with the development of depression in women than in men.13 (Data from the NCS also showed that among women, but not among men, between the ages of 15–54 years, having had an earlier psychiatric disorder was generally a stronger predictor of later alcohol dependence than was alcohol abuse, while lifetime co-occurrence of psychiatric and alcohol-related disorders was positively, but weakly, associated with the persistence of alcohol abuse among men and of alcohol dependence among both genders.14 ) In contrast, another study of lifetime mental disorders found that in all age groups depressed men were more
likely than depressed women to become problem drinkers.15 In a study based on Wave 1 data from the National Social Life, Health, and Aging Project (NSHAP),16 the authors of the current study found that heavy drinking was significantly associated with men’s depressive symptoms, but not with women’s.17 Drinking behavior and its correlates among older adults as well as gender differences in the relationship between alcohol use and mental health in middle and late adulthood are not well understood. The diverse and often conflicting findings of previous studies may stem from different samples and measures. In the present study, we used a large epidemiological dataset with a nationally representative sample of older adults to continue to explore the relationship between drinking and psychological distress in older adults by gender.
METHODS Data Source and Sample Data for this study are from the 2008 National Survey on Drug Use and Health (NSDUH).18 This annual survey provides information about the use of illicit drugs, alcohol, and tobacco among a nationally representative sample of members of the noninstitutionalized U.S. civilian population aged 12 or older. To improve the precision of the estimates, the sample allocation process targeted five age groups: 12–17, 18–25, 26–34, 35–49, and 50+. Data were collected using a combination of computer-assisted personal interviews and computer-assisted self-interviewing. The final total sample size for the 2008 NSDUH was 68,731, and the public use data file contains 55,739 cases. In the present study, we focused on 5,325 respondents who were aged 50 or older at the time of the survey. Measures Quantity and Frequency of Alcohol Consumption In the NSDUH, an alcoholic “drink” is defined as a can or bottle of beer, a glass of wine or a wine cooler, a shot of liquor, or a mixed drink with a shot of liquor in it. The quantity of alcohol used in the preceding 30 days was measured by two variables. First, the respondent was asked if he or she had consumed at least 5 drinks on a single occasion (ie, at the same time or within a 2-hour period) on at least 1 of the preceding 30 days. Regardless of the drinker’s gender, the NSDUH defines “binge” alcohol use as consumption of 5+ drinks on at least 1 of the preceding 30 days and “heavy” alcohol use as consumption of 5+ drinks on 5–30 of the preceding 30 days. Second, the respondent was also asked about the usual number of drinks per day he or she had consumed in the preceding 30 days. Since we wanted to identify whether a possible threshold—a specific number of drinks and/or a certain frequency of drinking—may present a significant risk factor for psychological distress, we decided to use Choi and DiNitto
categorical rather than continuous values in our analyses. The abstainer/nondrinker group included “never user” and “nondrinker in the preceding 30 days.” Based on a preliminary analysis, we categorized the number of drinks on a typical drinking day as follows: 1 drink, 2 drinks, 3 drinks, and 4+ drinks. We adopted the NSDUH’s definition of “binge” alcohol use as consumption of 5+ drinks on at least 1 of the preceding 30 days; however, we used the term “heavy” drinking to refer to consumption of 3+ drinks per drinking day. Frequency was measured by the number of the preceding 30 days on which the respondent had consumed any alcoholic beverage. Responses were categorized as abstainer/nondrinker (never user or nondrinker within the preceding 30 days): 1–2 days, 3–5 days, 6–19 days, and 20–30 days. The usual number of drinks per drinking day and the frequency of drinking during the 30-day period were weakly correlated (Kendall’s tau-b = .23 for men and .17 for women, with both groups at p < .000). In addition to the number of drinking days in the preceding 30 days, data on the number of drinking days in the preceding year were also collected. The responses were categorized as abstainer/nondrinker (never user or nonuser within the preceding 12 months): 1–11 days, 12–49 days, 50–99 days, 100–299 days, and 300–365 days.
Psychological Distress This was measured by the six-question, short-form screening scale for psychological distress originally developed for and included in the 1997 and 1998 U.S. National Health Interview Survey (NHIS). The scale, known as K6, was found to have strong psychometric properties and to strongly discriminate between community cases and noncases of DSM-IV disorders among general population groups in and outside the United States.19,20 K6 items are during the past 30 days, how often did you feel (1) nervous; (2) hopeless; (3) restless or fidgety; (4) so sad or depressed that nothing could cheer you up; (5) that everything was an effort; and (6) down on yourself, no good, or worthless? A 5-point response scale (4 = all the time; 3 = most of the time; 2 = some of the time; 1 = a little of the time; and 0 = none of the time) was used for each question, with a maximum possible score of 24. The Cronbach’s alpha for the study sample was .82. In the current study, the K6 score was used as a continuous variable in bivariate and multivariate analyses. We also estimated the prevalence of severe psychological distress among the study sample. A K6 score of 13 or higher is accepted as the optimal cut-off point for severe psychological distress.21 In addition to the 30-day K6 score, we also report the K6 score for the worst month in the preceding year for reference. Sociodemographics The NSDUH public-use data file contains only categorical values for age, education, and family income. September–October 2011
Thus, age groups were defined as 50–64 and 65+; education as less than high school, high school graduate, some college, and college graduate; and annual family income as under $20,000, $20,000–$49,999, $50,000–$74,999, and $75,000+. The other sociodemographic variables were gender, marital status (married, divorced/separated, widowed, and never married), employment status (currently working for pay and not working for pay), and race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, Asian, non-Hispanic multiracial, Native American/Alaskan Native, and Native Hawaiian or other Pacific Islander). Given the small numbers of Native Americans/Alaskan Natives (n = 22) and Native Hawaiians/other Pacific Islanders (n = 11), the two groups were combined as “Other” in the multivariate logistic regression analysis. As a result, caution is required for interpretations related to the “Other” racial category.
Health Status This was measured with the number of diagnosed chronic medical conditions (diabetes, heart problems, high blood pressure, and stroke; range 0–4) and self-ratings of health (1 = poor to 5 = excellent). History of Mental Health and Alcohol Treatment Respondents were asked if they had received any treatment (counseling and/or medication) for their feelings or moods in the preceding year. Respondents were also asked if they had ever received any treatment for problems related to their alcohol or other drug use. The response categories were yes ( = 1) or no ( = 0) for both questions. Although there was a separate question about alcohol/drug treatment in the preceding year, we chose lifetime treatment history rather than preceding year history due to the low positive response count for the latter. Analysis Methods We conducted bivariate analyses to identify gender similarities and differences in the study variables of interest. With respect to alcohol consumption, we examined gender differences in self-reported binge drinking in the preceding 30 days, usual number of drinks per day in the preceding 30 days, and drinking frequency in the preceding 30 days and 12 months. Chi-square tests or independent sample t-tests (with Bonferroni correction for multiple comparisons) were used to examine gender differences in drinking quantity and frequency between all men and all women, and then, between drinking men and drinking women only. Using chi-square and t-tests, we also examined whether binge drinkers differed from other drinkers of the same gender with respect to K6 scores, physical health status, age of first alcohol use, and history of mental health and alcohol treatment. Then, with the continuous 30-day K6 score as the dependent variable, we employed ordinary least squares (OLS) regression analysis for each gender group sepa422
Drinking, Distress, and Older Adults
rately. In Model 1, we entered the binge drinking variable (binge use vs. nonbinge use of alcohol during the preceding 30 days, with the latter as the reference category), controlling for sociodemographic characteristics, health status, and mental health and alcohol treatment variables. In Model 2, we entered the usual number of drinks in the preceding 30 days (with the nonuse of alcohol as the reference category) and the same control variables. In Model 3, we entered the number of drinking days in the preceding 30 days (with the nonuse of alcohol as the reference category) and the same controls.
RESULTS Sample Characteristics As Table 1 shows, women composed 53.8% of the total sample. The men and women did not differ significantly by race/ethnicity. Compared to the men, the women were less likely to be college graduates, to be in the highest income group, and to be working for pay. The women had fewer diagnosed chronic medical conditions, and they rated their health as being better, but they were more likely to be depressed and have received treatment for emotional problems in the preceding 12 months. The women were more likely to be lifetime abstainers, less likely to have used alcohol in the preceding 30 days, more likely to have initiated alcohol use at a later age, and less likely to have received treatment for an alcohol problem. Quantity and Frequency of Alcohol Consumption Table 2 describes the men’s and women’s drinking behaviors in greater detail. In the preceding month, the men were substantially more likely than the women to have engaged in binge drinking. Of the men, 21.8% had engaged in binge drinking (ie, 5 drinks on a single occasion) on at least 1 day and 7.2% had engaged in binge drinking on 5–30 days, compared to 7.0% and 1.5% of the women, respectively. Among the men who drank at all in the preceding month, 38.7% had engaged in binge drinking on at least 1 day and 12.9% had done so on 5–30 days. The comparable figures for women are 17.2% and 3.7%. Among those who drank in the preceding month, women were more likely than men to have consumed 1 drink on the days in which they drank; similar percentages of men and women consumed 2 drinks, and men were more likely than women to have consumed 3+ drinks. As expected, a significant proportion of heavy drinkers (77.7% for men and 61.6% for women) reported binge alcohol use. Among those who drank in the preceding month, women were more likely than men to have consumed alcohol 1–2 days and 3–5 days, and men were more likely than women to have consumed alcohol 6–19 days and 20–30 days. Both the mean and median numbers of drinking days were higher for men than for women. With regard to preceding year alcohol use, women were more likely than September–October 2011
TABLE 1. Sample characteristics
Variable n (%)
All 5,325 (100%)
Age group (%) 50–64 65+ Marital status (%) Married Widowed Divorced/separated Never married Race/ethnicity (%) Non-Hispanic White Non-Hispanic Black Hispanic Asian Non-Hispanic multiracial Native American/Alaskan Native Native Hawaiian/Pacific Islander Educational level (%)