Update on Alcohol and Health Current Evidence July-August

Update on Alcohol and Health: Current Evidence July-August 2006 www. alcoholandhealth. org 1

Studies on Assessments and Interventions www. alcoholandhealth. org 2

Combining Medication With Specialist Behavioral Intervention for Alcoholism: the COMBINE Study Anton RF, et al. JAMA. 2006; 295(17): 2003– 2017. www. alcoholandhealth. org 3

Objectives/Methods n n To study the effects of pharmacotherapy with behavioral intervention for alcohol dependence 1383 recently abstinent patients with alcoholism randomized to 1 of 9 treatments for 16 weeks: n n n Placebo; Naltrexone (100 mg per day), acamprosate (3 g per day), or both; or Combined behavioral intervention (CBI) alone, with active pills, or with placebo www. alcoholandhealth. org 4

Objectives/Methods (cont. ) n CBI: n n “Medical management” (MM) provided to all but CBI alone group n n Offered by a specialist Included elements of cognitive behavioral therapy, motivational interviewing, and 12 -step facilitation Up to 20 50 -minute sessions 9 counseling and education sessions (45 -minute initial session and ≈20 -minute follow-up sessions) Provided by a generalist healthcare professional Focused on medication side effects, adherence, and alcohol abstinence 94% follow-up at 16 weeks; 82% a year later www. alcoholandhealth. org 5

Results n Drinking outcomes improved substantially in all groups. n The range of mean percent days abstinent across groups was… n 23%– 30% at baseline, n 67%– 81% at 16 weeks, and n 59%– 69% a year later. www. alcoholandhealth. org 6

Results (cont. ) Key Findings at 16 Weeks Naltrexone/MM n. Better than placebo/MM at increasing percent days abstinent (81% vs 75%) and delaying a return to heavy drinking* (hazard ratio 0. 7) n. Adding CBI/MM with any pills n. Better CBI alone (without MM or pills) n. Less Acamprosate/MM n. Did CBI did not improve these outcomes than placebo/MM at increasing percent days abstinent (e. g. , 79% versus 75%) effective than placebo/MM at increasing days abstinent (67% versus 74%) not significantly affect drinking outcomes *>=5 standard drinks per day for men, >=4 for women www. alcoholandhealth. org 7

Results (cont. ) n One year after treatment… n drinking outcomes did not significantly differ among groups. www. alcoholandhealth. org 8

Conclusions/Comments n This large, rigorous, and complex study showed that. . . n naltrexone and specialist counseling have similar modest efficacy when each is offered with MM, and specialist counseling alone has less efficacy than when combined with MM. These findings support the use of naltrexone and intensive MM in primary care settings as a potential alternative to specialized treatment. www. alcoholandhealth. org 9

Conclusions/Comments (cont. ) n n Delivering the type of medical management provided in this study requires substantial training and a collaborative care model. This requirement and other concerns (e. g. , need for chronic treatment, acamprosate’s perplexing lack of efficacy) require further investigation. www. alcoholandhealth. org 10

Acamprosate May Work Only When Abstinence Is the Goal Mason BJ, et al. J Psychiatr Res. 2006; 40(5): 382– 392. www. alcoholandhealth. org 11

Objectives/Methods n To evaluate acamprosate’s efficacy in U. S. patients n Randomized controlled trial of the drug (the first in this country) in 21 alcohol treatment clinics n Subjects assigned to either… n n the standard 2 g of acamprosate per day (n=258), 3 g per day (n=83), or placebo (n=260) All received self-help materials and 8 sessions of brief counseling www. alcoholandhealth. org 12

Results n n The percentage of days abstinent at 6 months did not significantly differ across groups in unadjusted analyses. However, in analyses adjusted for potential confounders (e. g. , readiness to change, treatment goal of abstinence), the percentage was… n significantly higher in subjects who received acamprosate: n 52% for placebo, 58% for subjects on 2 g, and 63% for subjects on 3 g. www. alcoholandhealth. org 13

Results (cont. ) n Patients who stated at baseline that abstinence was their goal achieved the best results: 58% days abstinent for placebo n 70% for subjects on 2 g n 73% for subjects on 3 g n www. alcoholandhealth. org 14

Comments/Conclusions n n n Acamprosate’s lack of efficacy in this large, well-designed trial is surprising, though also noted in another recent study. * The U. S. subjects may have been abstinent for a shorter time than were subjects in international studies showing acamprosate’s efficacy. The link between having abstinence as a goal and benefit from acamprosate may guide clinicians’ prescribing practices and merits confirmation. *Anton RF, et al. JAMA. 2006; 295(17): 2003– 2017. www. alcoholandhealth. org 15

Most Medical Inpatients With Unhealthy Alcohol Use Have Alcohol Dependence Saitz R, et al. J Gen Intern Med. 2006; 21(4): 381– 385. www. alcoholandhealth. org 16

Objectives/Methods n n n To determine the prevalence and severity of unhealthy alcohol use on a medicine service 5813 medical inpatients screened for drinking risky amounts* 341 who drank risky amounts were evaluated further *>14 standard drinks per week or >=5 drinks per occasion for men (>11 and >=4, respectively, for both women and people >=66 years) www. alcoholandhealth. org 17

Results n 17% of inpatients screened drank risky amounts. n Of those drinking risky amounts… n n n 97% exceeded per occasion limits, * and most scored >=8 on the Alcohol Use Disorders Identification Test (strongly correlating with a current alcohol diagnosis). Of the 341 inpatients who drank risky amounts and received more detailed evaluation… n 77% had alcohol dependence. *>=5 drinks for men, >=4 for women and people >=66 years www. alcoholandhealth. org 18

Conclusions/Comments n According to this study, almost 1 in 5 medical inpatients has unhealthy alcohol use… n n a number identical to that found in a recent systematic review of hospital alcohol screening studies. * Of note, most medical inpatients with unhealthy alcohol use have alcohol dependence. *Roche AM, et al. Drug Alcohol Depend. 2006; 83(1): 1– 14. www. alcoholandhealth. org 19

Conclusions/Comments (cont. ) n Because brief intervention—the currently recommended practice—has established efficacy only for nondependent unhealthy alcohol use… n new strategies to address alcohol dependence on a medicine service are warranted. www. alcoholandhealth. org 20

Another Single-Item Screening Test? O’Brien MC, et al. Acad Emerg Med. 2006; 13(6): 629– 636. www. alcoholandhealth. org 21

Objectives/Methods n n To examine whether one question (“In a typical week, how many days do you get drunk? ”) could identify drinkers at risk of injury Web-based survey completed by 3909 college students in North Carolina n 36% response rate www. alcoholandhealth. org 22

Results n Of 2488 current (past 30 -day) drinkers… n n 54% got drunk at least once in a typical week. Drunkenness was a better indicator of injury outcomes than was heavy episodic drinking. * *>=5 drinks in a row on at least 1 day in the past 30 days for men, >=4 for women www. alcoholandhealth. org 23

Results (cont. ) Odds of Injury in Students Who Had Been Drunk in a Typical Week Vs. Those Who Had Not Injured because of their drinking OR* 5. 0 Had a fall that required treatment OR 2. 2 Taken advantage of sexually as a result of another’s drinking OR 2. 6 Caused injury to another that required treatment OR 2. 6 *Odds ratios adjusted for potential confounders (e. g. , race, current drinking) www. alcoholandhealth. org 24

Conclusions/Comments n n n This study did not assess for alcohol use disorders, a significant limitation that precludes recommending the single question as a clinical screening test. But the results are interesting because they suggest that… n a question about drunkenness might identify collegestudent drinking with consequences better than a question about heavy drinking on an occasion would. For now, it seems reasonable for clinicians to ask the question when discussing drinking with college students. www. alcoholandhealth. org 25

Intensive Referral to 12 -Step Groups Improves Outcomes Timko C, et al. Addiction. 2006; 101(5): 678– 688. www. alcoholandhealth. org 26

Objectives/Methods n To assess whether intensive referral is more effective than standard referral at… n n n increasing involvement with Alcoholics Anonymous (AA) and Narcotics Anonymous (NA), and subsequently improving substance use outcomes Study of 345 veterans entering outpatient substance abuse treatment n n Almost all had attended 12 -step meetings previously 46% preferred alcohol to other substances www. alcoholandhealth. org 27

Objectives/Methods (cont. ) n Subjects were randomized to either… n n standard referral, including a schedule of and encouragement to attend local 12 -step meetings, or intensive referral, including additional information and clinician support (e. g. , linkage to AA/NA volunteers, follow-up on meeting attendance, encouragement to obtain a sponsor) www. alcoholandhealth. org 28

Results n n n At 6 months, attendance at 12 -step meetings did not significantly differ among the groups. But, of subjects with less exposure to 12 -step groups at study entry, those assigned to intensive referral had better attendance. The intensive referral group had significantly greater… n 12 -step involvement (e. g. , obtained a sponsor) and improvement in substance use problems. www. alcoholandhealth. org 29

Results (cont. ) n The intensive referral group was also more likely to be abstinent from… drugs (78% versus 70%, P<0. 05) and n alcohol (76% versus 70%, P<0. 12). n n Involvement in 12 -step groups partially mediated improvements in alcohol outcomes. www. alcoholandhealth. org 30

Conclusions/Comments n n n Because 12 -step groups do not sponsor research, rigorous trials of their effectiveness are difficult to perform. This study provides strong evidence that clinicians should support patients’ participation in AA/NA. Linking patients with AA/NA volunteers, following up on attendance, and encouraging sponsorship are essential to successful 12 -step facilitation. www. alcoholandhealth. org 31

Help Seeking Quadruples the Likelihood of Abstinence Dawson DA, et al. Addiction. 2006; 101(6): 824– 834. www. alcoholandhealth. org 32

Objectives/Methods n n To quantify the effect of help seeking on recovery from alcoholism Analysis of data from 4422 adults in the U. S. who had… participated in a nationally representative survey and n developed alcohol dependence at least 1 year before their participation n www. alcoholandhealth. org 33

Results n Only 26% of subjects had ever sought help for their alcohol problems: n n 3% participated in a 12 -step program only, 6% in formal treatment only, and 17% in both. Help seekers drank more than subjects who had not sought help and had higher lifetime prevalences of… n other drug use, mood disorders, and personality disorders. www. alcoholandhealth. org 34

Results (cont. ) n In analyses adjusted for potential confounders, help seeking significantly increased the likelihood of… n n n any recovery* (OR 2. 4) and abstinence (OR 4. 0). The odds of recovery were greater for those who had… n participated in 12 -step programs with or without formal treatment than for those who had participated in formal treatment only. *In the past year, having no symptoms of alcohol abuse or dependence and either drinking low-risk amounts (<=14 drinks per week and <=4 drinks on any day for men; <=7 drinks per week and <=3 drinks on any day for women) or abstaining www. alcoholandhealth. org 35

Conclusions/Comments n n n Even though they had more comorbidity, seekers of formal and informal treatment had better odds of recovery from alcohol dependence. This study could not separate the motivation inherent in seeking help from therapeutic effects of help received. However, help seeking—regardless of the patient’s level of readiness—should be encouraged. www. alcoholandhealth. org 36

Should Screening for Colorectal Cancer Start Earlier for Drinkers and Smokers? Zisman AL, et al. Arch Intern Med. 2006; 166(6): 629– 634. www. alcoholandhealth. org 37

Objectives/Methods n n n To explore whether earlier screening for colorectal cancer (CRC) is warranted in patients who smoke and drink 166, 172 cases of CRC identified through a national medical registry; age at diagnosis assessed Analyses controlled for sex, race, and insurance status www. alcoholandhealth. org 38

Results n n n CRC was diagnosed significantly earlier in current and past drinkers* than in subjects who never drank. Diagnosis occurred even earlier among current and past drinkers who currently smoked (e. g. , 7. 8 years earlier in current drinkers who also smoked than in subjects who never drank or smoked). The likelihood of distal CRC was significantly higher among… n n current and past drinkers than in subjects who never drank (ORs 1. 2 and 1. 1, respectively) and current smokers than in subjects who never smoked (OR 1. 2). *Current use defined as use in the past year; past use defined as having quit at least 1 year before CRC diagnosis www. alcoholandhealth. org 39

Conclusions/Comments n n This large retrospective study found that CRC was diagnosed in current and past drinkers at a younger age, especially if they currently smoked. While the study helps clarify the association between alcohol, tobacco, and CRC, it… n neither addressed whether screening could have detected the earlier diagnoses n nor accounted for many important confounding factors (e. g. , quantity/frequency of drinking and smoking, diet, comorbidities that potentially lead to earlier gastrointestinal evaluation). www. alcoholandhealth. org 40

Conclusions/Comments (cont. ) n Thus, more investigation is needed to determine whether… n CRC screening guidelines should be modified according to a patient’s history of alcohol and tobacco use. www. alcoholandhealth. org 41

Studies on Alcohol and Health Outcomes www. alcoholandhealth. org 42

Is Drinking Unsafe for Patients on Warfarin or Statins? Mukamal KJ, et al. Am J Med. 2006; 119(5): 434– 440. www. alcoholandhealth. org 43

Objectives/Methods n n To assess drinking’s safety in warfarin or statin users 1244 men with previous coronary artery bypass graft surgery who had enrolled in a randomized trial of daily… n n lovastatin (mean dose of 4 mg or 76 mg), low-dose warfarin (1– 4 mg to achieve an INR* of 1. 8– 2), or placebo-warfarin 54% drank <1 drink per week; 9 men drank >21 drinks per week *International normalized ratio www. alcoholandhealth. org 44

Results n During about 5 years of follow-up, alcohol use did not significantly affect the risk of having elevated INR or ALT* levels: Standard Drinks per Week INR>=2 (% of (warfarin subjects only) <1 8 67 1– 6 10 66 7– 13 9 68 >=14 § ALT>=80 IU/L (% of subjects) 6 61 ALT results were similar when the analysis was restricted to patients taking the higher dose of lovastatin. *Alanine aminotransferase www. alcoholandhealth. org 45

Conclusions/Comments n Few men drank >21 drinks per week, so this study could not address the risks associated with warfarin or statin use and heavy drinking. n Reporting abnormal creatine kinase levels by drinking categories would have helped readers to judge risk. n These results should somewhat reassure patients with coronary artery disease that… n drinking moderately while taking warfarin or lovastatin is not harmful (or at least does not increase the risk of developing 2 specific lab abnormalities). www. alcoholandhealth. org 46

Alcohol and the Risk of Injury Gmel G, et al. Alcohol Clin Exp Res. 2006; 30(3): 501– 509. www. alcoholandhealth. org 47

Objectives/Methods n n n To examine whether specific patterns of alcohol use affect the risk of injury Assessment of 8736 patients admitted to an emergency department in Switzerland (5077 with an injury) Measures: n n n Usual volume of drinking Past-month heavy episodic drinking (>=5 drinks on at least one occasion for men, >=4 for women) Recent drinking (in the 24 hours before the emergencydepartment visit) www. alcoholandhealth. org 48

Results n Heavy episodic drinking and recent drinking increased the risk of injury. n As the volume of usual and recent drinking increased, the risk of injury increased. n Risk was highest in patients who… n usually drank moderately* and reported both pastmonth and recent heavy episodic drinking. n Odds ratios (ORs) 6. 4 for men and 7. 4 for women, compared with abstainers *<14 drinks per week for men, <7 for women www. alcoholandhealth. org 49

Results (cont. ) n n Past-month heavy episodic drinking tended to confer a higher risk in patients who usually drank moderately than in patients who usually drank heavily. Almost half of the alcohol-attributable injuries among women were suffered by those… n n n who usually drank moderately, without past-month heavy episodic drinking, and drank >0 but <4 drinks in the 24 hours before their emergency-department visit. www. alcoholandhealth. org 50

Conclusions/Comments n n This study indicates that patterns of drinking, particularly heavy episodic drinking, influence the risk of injury. Interventions to reduce alcohol-related injury should focus on… n preventing heavy episodic drinking among both moderate and heavy drinkers. www. alcoholandhealth. org 51

Lifetime Drinking, Confounders, and Breast Cancer Risk Terry MB, et al. Ann Epidemiol. 2006; 16(3): 230– 240. www. alcoholandhealth. org 52

Objectives/Methods n To examine whether confounding factors can explain the association between moderate drinking and an increased risk of breast cancer n Data from a population-based study n Comparison of 1508 women with breast cancer and 1556 matched controls n Analyses adjusted for potential confounders www. alcoholandhealth. org 53

Results n Risk was not significantly associated with lifetime* or current alcohol use, except for current drinking of <0. 5 grams per day (OR 0. 7). Results from Subgroup Analyses Subgroup Breast Cancer Risk Associated With Lifetime Intake of 15 -30 g of Alcohol per Day BMI <25 OR** 2. 1 Post-menopausal and never used hormone replacement therapy OR 2. 0 *Calculation based on average intake during various periods across the lifetime **Versus nondrinking www. alcoholandhealth. org 54

Results (cont. ) n Lifetime drinking of >=15 g per day was significantly associated with… estrogen receptor positive tumors (OR 1. 9) and invasive tumors (OR 1. 6), n but only in women with a BMI <25. n n Results did not differ in analyses that examined alcohol intake during a particular decade of life. www. alcoholandhealth. org 55

Conclusions/Comments n n n These findings are consistent with those of most previous studies that show a weak positive relationship between alcohol use and breast cancer risk. There was no dose-response curve for any of the observed associations, and only the subgroup analyses were significant. Additional studies are needed to help clinicians identify women at higher risk of breast cancer from current or lifetime moderate drinking. www. alcoholandhealth. org 56

Discrimination Against People With Alcoholism Schomerus G, et al. Drug Alcohol Depend. 2006; 82(3): 204– 210. www. alcoholandhealth. org 57

Objectives/Methods n To examine whether public attitudes and beliefs about alcohol dependence… n n are more negative than those about other diseases and affect public preferences for resource allocation Other diseases = schizophrenia, depression, Alzheimer’s disease, rheumatism, diabetes, AIDS, myocardial infarction, and cancer Phone interviews of 1012 German adults www. alcoholandhealth. org 58

Results Respondents… % of Respondents With Belief/Attitude Would spare alcohol treatment from budget cuts if resources were scarce 7% Would distance themselves from people with alcoholism 78% (moreso than any other disease) Believed alcoholism is self-inflicted 85% (moreso than any other disease) Believed alcoholism is severe 55% Thought alcoholism can be treated effectively 30% Felt at risk for alcoholism 4% www. alcoholandhealth. org 59

Results (cont. ) n Respondents were more likely to choose budget cuts for alcohol treatment if they believed that… n n n alcoholism was self-inflicted, alcoholism was not as severe as other conditions, and their risk of alcoholism was low. www. alcoholandhealth. org 60

Conclusions/Comments n This study highlights… the public’s lack of understanding about alcohol dependence and n the severe stigma faced by people with the disease. n n Increasing awareness and destigmatizing alcoholism are essential to improve the delivery of quality care. www. alcoholandhealth. org 61

Lifetime Drinking and the Metabolic Syndrome Fan AZ, et al. Euro J Epidemiol. 2006; 21(2): 129– 138. www. alcoholandhealth. org 62

Objectives/Methods n To determine whether lifetime drinking affects the risk of the metabolic syndrome n Population-based sample of 2818 subjects aged 35– 79 years n n Subjects drank >=1 a month for >=6 months during their lifetimes and had no cardiovascular disease or cancer when they were interviewed Analyses adjusted for potential confounders www. alcoholandhealth. org 63

Results As lifetime drinking intensity* increased… n the prevalence of the metabolic syndrome and most of its components significantly increased. Quartile of Lifetime Drinking Intensity Prevalence Ratios of Metabolic Syndrome (vs. subjects in the lowest quartile) Second 1. 2 Third 1. 4 Fourth 1. 6 *Total number of drinks over a lifetime divided by the total number of drinking days over a lifetime www. alcoholandhealth. org 64

Results n As the number of lifetime drinking days increased… n the prevalence of abdominal obesity in women and low high-density lipoprotein cholesterol significantly decreased. www. alcoholandhealth. org 65

Conclusions/Comments n Studies of lifetime drinking are desirable, but their validity can be difficult to judge because they are rarely prospective. n n n Results depend on the recall of amounts consumed in the distant past; any errors in these estimates get multiplied when calculating lifetime intake. But, at least in this study, the researchers used an extensively studied, intensive interview method with known reliability. Further, the results support past findings: increasing peroccasion amounts of alcohol are harmful, while more frequent drinking may be protective. www. alcoholandhealth. org 66