Alcohol use declines with age. In the 2001–2002 NESARC, 45.1% of the participants aged 65+ reported that they had consumed one or more drinks in the past year, compared to 64.3% among those 45–64, 72.9% among those 25–44 and 70.8% among those 18–24. Among those aged 65+, 28.5% reported former alcohol use and 26.4% reported lifetime abstinence. Among older adults, the prevalence of current drinkers was 48.3% among Whites, 23.4% among Blacks, 37.9% among American Indians/Alaskan Natives, 32.7% among Asian/Pacific Islanders, and 36.6% among Hispanics. Among current drinkers aged 65+, 64.3% were classified as light drinkers, 20.7% as moderate drinkers and 13.9% as heavy drinkers. A total of 58.0% of older adults with excellent/very good self-perceived health were current drinkers compared to 29.1% of those who rated their health as fair/poor¹³.

This decline in alcohol use by age is also observed among the oldest old. Ruchlin examined the prevalence of alcohol consumption in adults aged 55+ and found 46% had consumed alcohol in the past year, with a continuous decline across age groups, from 52.9% of those aged 55–64 categorized as current drinkers compared to 24.7% of those aged 85+. A total of 17% of the sample reported that they had consumed alcohol every day in the past two weeks. In controlled regression analyses, more people aged 65–74 drank every day, compared to those 55–64 (odds ratio = 1.36), but people aged 75+ drank less than those aged 55–64. Males and Whites used alcohol more frequently and were more likely to be heavy drinkers than females and non-Whites. The lower one’s perceived health status, the lower the odds of drinking every day. Believing excessive drinking increased the chances of getting cirrhosis of the liver decreased the odds of moderate and heavy drinking¹⁴.

There are various explanations as to why the prevalence of alcohol use is lower in older adults. Selective survival may be a factor, in that persons who drink may be less likely to survive to older ages. Cohort effects are also possible. Persons who grew up before the Second World War may have had lower alcohol use throughout their lives. In the NESARC, the proportion of lifetime abstainers was highest in the 65+ age group (26.4 vs. 13.7% among those aged 25–44 years)¹³. Studies have also shown that some older adults decrease their use as they grow older. Busby et al. investigated alcohol use in a community-based sample of adults aged 70+ in New Zealand. Both frequency and quantity of intake decreased with age. A total of 60.1% of the men and 30.3% of the women said they drank less compared to middle age, while 7.4% of the men and 11.1% of the women said they drank more. The main reasons cited for decreased use of alcohol were change in health and fewer social opportunities, with reasons cited for increased intake being more time and money¹⁵.

Adams et al. followed a cohort of 270 healthy community-dwelling adults aged 60+ over seven years. At baseline, the investigators found a decline in percentage of drinkers with increasing age. In the seven-year follow-up, there was a 2% per year decline in the percentage of subjects consuming any alcohol, but the mean alcohol intake did not change for those who continued to drink, except among heavy drinkers, suggesting an age-related decline rather than a cohort effect¹⁶. A study from Denmark of adults aged 50–74 interviewed at five time points between 1987 and 2005 reported that the unadjusted probability of heavy drinking declined with age but increased by calendar year and year of birth for both men and women¹⁷.

Using data from four time points over 20 years, Karlamangla et al. found evidence for age and period effects for heavy drinking but not cohort effects across all age groups. The prevalence of heavy drinking decreased with increasing age, an age effect, and also tracked declines in USA per capita alcohol consumption, a period effect. The prevalence of heavy drinking in specified age groups did not differ by birth year¹⁸. Cohort effects have been noted across all age groups, however, for beverage-specific consumption, with adults born before 1940 more likely to drink spirits than those born after 1940, and elevated beer consumption among those born between 1946 and 1965¹⁹. In the NESARC, beverage preference among current drinkers aged 65+ was coolers (long drinks with ice, 1.3%), beer (23.6%), wine (31.1%), liquor (14.2%) and no preference (29.7%). Preference for beer was lower and preference for wine and liquor was higher compared to other age groups¹³.

While the NESARC reported that less than half of older adults in the USA had consumed any alcohol in the previous year, regional studies have shown much variation. In a sample of 270 healthy men and women aged 65–89 living in the Southwestern USA, 48% of the participants reported in their three-day diet record that they had consumed alcohol, with 66% reporting they had consumed alcohol at least monthly²⁰. The prevalence of alcohol use reported from the Established Populations for Epidemiologic Studies of the Elderly (EPESE) studies of persons aged 65+ varied by site. In East Boston, 70.5% of the sample drank alcohol in the past year and 54.7% had used alcohol in the past month. Similar findings were reported from New Haven, where 65.8% had used alcohol in the past year and 51.9% in the past month. The proportions were lower in Iowa and North Carolina. In the Iowa sample, 46.3% had used alcohol in the past year, while 31.2% had done so in the past month. In the North Carolina EPESE, 37.4% had consumed alcohol in the past year and 24.6% in the past month^21,22.

Samples from clinical populations have also found a range of alcohol use among older adults. Saunders et al. reported, among men and women aged 65+ randomly selected from patient rosters, that 10.5% admitted to drinking more or less every day (17.7% men and 6.1% women). At the three-year follow-up, one-fifth of the participants were regular drinkers, drinking at least on one occasion per week²³. In a sample of 539 medical admissions aged 65+, the prevalence of alcohol abuse was 7.8%. An additional 29.7% of the sample who were neither abstainers nor occasional drinkers or alcohol abusers drank regularly: 43% of the men and 16% of the women²⁴.

Heavy drinking in older adults has been documented in research studies in various ways. In the EPESE baseline studies, the percentage of persons who drank two or more ounces of absolute alcohol per day in the previous month was 8.4% in East Boston, 6.6% in New Haven, 5.4% in Iowa and 7.2% in North Carolina^21,22. Goodwin et al. found in their sample from the Southwestern USA that 17% of the drinkers drank more than 30 g alcohol per day on average²⁰. In a study of older Medicare beneficiaries, 9% reported exceeding recommended drinking limits. The prevalence was higher in men (16%) than women (4%). Exceeding limits was defined as monthly use exceeding 30 drinks per typical month, or consuming four or more drinks a day in a typical month²⁵. In a report from the 2004 Behavioral Risk Factor Surveillance System (BRFSS), the prevalence of binge drinking decreased with age, from 27.4% among those aged 18–24 to 3.7% among those aged 65+. However, among binge drinkers, those aged 65+ reported the highest average number of binge drinking episodes during the previous month. The number of drinks consumed during the most recent binge episode decreased with age, from 9.8 among those aged 18–24 to 6.4 among those aged 65+²⁶.

The prevalence of heavy alcohol use among drinkers has also been reported. In the NESARC, daily limits across all age groups were defined as 5+ drinks per day for men and 4+ for women. A total of 3.9% of older drinkers reported excess drinking 1–11 times in the previous year, and 6.1% reported excess consumption 12+ times in the past year. Weekly limits were defined as 14 drinks per week for men and 7 per week for women. A total of 13.9% of the older drinkers exceeded the weekly or both daily and weekly limits during the previous year¹³. Saunders et al. found that 19.5% of the men and 19.6% of the women who were regular drinkers were exceeding sensible limits²³.

The prevalence of heavy drinking in clinical samples is similar. Adams et al. screened 5065 primary care patients aged 60+ and found that 15% of the men and 12% of the women regularly drank in excess; 9% of the men and 2% of the women reported regularly consuming more than 21 drinks per week²⁷. Iliffe et al. studied 241 patients aged 75+ from general practice and found that 51% of the men and 22% of the women reported using alcohol in the past three months. Among drinkers, 3.6% of the men admitted consuming more than 21 units and 3.2% of the women more than 14 units of alcohol per week. Weekly alcohol consumption was not associated with age, cognitive impairment, depression, falls or inpatient or outpatient care²⁸. Callahan and Tierney found the prevalence of CAGE²⁹-defined alcoholism was 10.6% among primary care patients aged 60+. Patients with alcoholism were more likely to be younger, have fewer years of education, and be male, Black, smokers and malnourished³⁰. In another sample of adults aged 65–103 years selected from primary care, 70% reported no alcohol consumption during the past year. A total of 21.5% were moderate drinkers, 4.1% were at-risk drinkers and 4.5% were heavy or binge drinkers³¹.

Bristow and Clare interviewed 650 medical and geriatric admissions over age 65 and found 9% of the men but few of the women drank in excess of recommended safety limits. Another 10% had drunk heavily over the age of 65 but cut down because of medical, social, financial or other reasons. Compared to the non-drinkers and light drinkers, the heavy drinkers were more likely to smoke, to be unmarried and to have some impairment of mobility³². In a similar sample of hospital patients aged 65+, 54% reported no alcohol consumption and 9% screened positive for alcohol problems (17% among male and 2.5% among female patients). Seven percent of the patients were discharged with an alcohol-related diagnosis³³.

Clinicians may have difficulty detecting and diagnosing alcohol problems in older adults for a number of reasons. Screening instruments used in younger populations to detect problem drinking may not be as reliable for older adults^34,35. Adams et al. compared responses to a beverage-specific self-administered questionnaire about the quantity and frequency of alcohol use and episodes of binge drinking to the widely used CAGE questionnaire (Cut down, Annoyed by criticism, Guilty about drinking, Eye-opener drinks)²⁹ in 5065 primary care patients age 60+, and found the CAGE performed poorly in detecting heavy or binge drinkers²⁷. Many of the screening instruments enquire about frequency and quantity of alcohol use, but even small amounts may be problematic for older adults with chronic health conditions and in combination with certain medications. Criteria for alcoholism often include problems with social and/or occupational functioning. However, older adults are less likely to be married or employed, and therefore less likely to report marital or job problems. Older drinkers may also be less likely to cause public disturbances resulting in legal problems³⁶.

Clinicians may be reluctant to enquire about alcohol use out of embarrassment or because the patient doesn’t fit the stereotype of an alcoholic. They may also fail to diagnose alcohol problems because symptoms of alcohol misuse may be similar to symptoms of ageing such as cognitive difficulties³⁷. Adams et al. screened patients 65+ seen in the emergency department for alcoholism. Using their criteria of either CAGE-positive or self-reported drinking problems and alcohol use within the past year, they found the current prevalence of alcohol abuse was 14%, with a high prevalence (22%) among those presenting with gastrointestinal problems. Physicians, however, detected only 21% of current alcohol abusers³⁸. In a study from primary care, only 41.6% of older patients who screened positive for alcoholism had a diagnosis of alcoholism recorded in their medical record³⁰. In a study where all new admissions to the medical service were screened for alcoholism using the CAGE and the Short Michigan Alcohol Screening Test (SMAST)³⁹, the prevalence of alcoholism was 27% in patents under 60 and 21% among patients aged 60+. These age differences were not significant. However, 60% of screen-positive younger patients were identified as having alcoholism by their house officers compared to only 37% of those aged 60+. Older patients with alcoholism were less likely to be diagnosed if they were White, female, or had completed high school. Even when diagnosed, older patients with alcoholism were less likely than younger patients to be referred for treatment⁴⁰.

Another identification problem is establishing what constitutes excessive drinking in older adults. Moos et al. compared several guidelines defining excessive drinking in a sample of 1291 older adults. Depending on the criteria used, 23 to 50% of the women and 29 to 45% of the men engaged in potentially unsafe drinking patterns. The number of drinks per week and per day was associated with alcohol use problems. The authors concluded that safe alcohol consumption levels should not necessarily be higher for men⁴¹