Definitional problems
A plethora of studies have estimated the incidence and prevalence of drug abuse in younger populations. However, far fewer studies have focused on elderly persons. Furthermore, several methodological problems, including inappropriate definitions, may have resulted in low estimates. Studies using currently available diagnostic criteria for substance abuse are likely to significantly underestimate the prevalence of drug abuse among elderly persons because the criteria were developed and validated in young and middle-aged samples; the criteria may have only limited utility among elderly populations (
3,
4,
5).
For example,
DSM-IV criteria (
6) include increased tolerance of the effects of the substance, which results in increased consumption over time. However, changes in pharmacokinetics and physiology may alter drug tolerance in elderly persons. Decreased tolerance of alcohol among older individuals may lead to decreased consumption with no apparent reduction in intoxication.
Another
DSM-IV criterion (
6) for abuse involves adverse consequences associated with substance use such as absences from work or poor work performance, suspensions or expulsions from school, and neglect of children. These consequences fail to consider age-related differences. One would expect fewer such adverse consequences to occur among elderly persons since they often live alone and are unemployed.
In contrast to younger substance abusers who most often abuse illicit drugs, substance abuse problems among elderly individuals may result from misuse of over-the-counter and prescription drugs. Misuse of drugs refers to underuse, overuse, or erratic use of legal drugs, either prescribed or over-the-counter drugs. In its extreme form, misuse may become drug abuse (
3,
6).
Prescription drugs
Elderly persons use prescription medications approximately three times as frequently as the general population (
7), and use of over-the-counter medications by this group is even more extensive (
8). In the United States, the estimated annual expenditure on prescription drugs by elderly persons is $15 billion—a four-fold-greater expenditure per capita on medications than that of younger individuals (
9,
10).
Psychoactive drug use is particularly problematic. Data obtained from elderly persons living in the community who were receiving services at a mental health clinic revealed that prescription drug abuse, most often of sedative-hypnotic, antianxiety, and analgesic drugs, accounted for about 5 percent of the average caseload (
11). Beers and associates (
12) reported that more than half of all residents of intermediate care facilities in Massachusetts were receiving psychoactive drugs, and 30 percent received long-acting drugs not recommended for elderly persons.
Other data from the Veterans Affairs hospital system suggested that inappropriately high doses of benzodiazepines were commonly prescribed for elderly patients (
13). A national survey of approximately 3,000 persons living in the community found that 1.6 percent had taken benzodiazepines daily for one year or longer and that older persons were overrepresented among the users; 71 percent of this group were over 50 years old (
14). Other estimates of psychoactive drug use in the elderly population range from 23 percent in the National Medical Care Expenditure Survey (
15) to 28 percent in an urban Seattle housing project (
16).
Comorbid disorders and gender appear to be important predictors of prescription drug use. Finlayson and Davis (
17) examined prescription drug use among 100 elderly patients admitted to an inpatient addiction program during a 20-year period (1974-1993) and found that 72 percent had an alcohol use disorder only, 16 percent had prescription drug dependence, and 12 percent had both alcohol and drug dependence. In this sample 35 percent developed drug dependence after age 60. The greatest risk factor for abuse of prescription drugs was being a woman. This finding is supported by other studies suggesting that elderly women are more likely to visit physicians and to receive prescriptions for psychoactive drugs than elderly men (
18,
19,
20,
21).
In contrast, Robins and Clayton (
22) analyzed data from the National Household Survey on Drug Abuse and concluded that older men were more likely than women to report use of sedatives, tranquilizers, and stimulants. Moos and associates (
23) examined inpatient treatment records for 1987 from Department of Veterans Affairs medical centers. A total of 98,000 patients had a diagnosis of substance abuse, and approximately 22 percent of them (21,139 persons) were age 55 or older. Of these, 13.7 percent were diagnosed as having drug dependence or drug-induced psychosis, and an additional 58.2 percent had a diagnosis of alcohol abuse or dependence.
Swartz and colleagues (
24) examined wave 1 data from the Piedmont health survey, which was part of the Epidemiologic Catchment Area (ECA) study, and reported that benzodiazepine use was predicted by being elderly, white, female, less educated, and separated or divorced, by having experienced a greater number of negative life events, and by having a psychiatric diagnosis.
An association may exist between age-related physical morbidity and abuse of medications. For example, individuals with arthritis may grow increasingly dependent on pain medications, and those with sleep problems may be more likely to abuse benzodiazepines. This association may be partly due to difficulties that older individuals have in following and reading prescriptions (
25). In addition, elderly patients are more likely to be prescribed medications for longer periods of time than younger patients (
26).
Illegal drug use
Cross-sectional data support the findings of a low prevalence rate of illicit drug use in the elderly population. Less than .1 percent of individuals older than 65 in the ECA study met
DSM-III criteria (
27) for drug abuse or dependence during the previous month (
1). The prevalence rate was 3.5 percent for the same period among 18- to 24-year-old persons. ECA data suggest a lifetime prevalence of illegal drug use of 1.6 percent for persons over age 65 (
28).
In addition, data gathered through other means—for example, surveys of homeless individuals, reports of adverse drug reactions, and drug arrests—indicate that illicit drug use in the elderly population is very uncommon (
29). In 1982 the one-year prevalence of marijuana use among persons over age 50 was 1 percent (
29). Data from the Drug Abuse Warning network in 1991 indicated that 1.8 percent of emergency department contacts for heroin or morphine abuse were for persons over age 55 (
30). Approximately 2 percent of all methadone maintenance clients in New York City in 1985 were over age 60 (
31).
Atkinson and colleagues (
29) reported that development of addiction after young adulthood is rare and that mortality among addicted individuals is high. For example, in one 24-year follow-up study of heroin addicts, more than 27 percent of subjects died during the study period (
32). National data indicate that 5.6 percent of deaths associated with heroin or morphine use were of persons older than 55 (
33). Solomon and Stark (
34) reported that only two of 26 elderly men in a substance abuse program had ever used illicit drugs, suggesting that drug abuse in that age group largely involves prescription and over-the-counter drugs and alcohol.
Few longitudinal studies that allow examination of drug use trends over time in specific age cohorts have been carried out. The National Survey on Drug Abuse is an annual nationwide study of the household population of the United States that is sponsored by the National Institute on Drug Abuse and has been conducted since 1971 (
35). Although published data group together respondents over age 35, it is possible to obtain a rough estimate of the impact of aging, at least into middle age, and the aging of baby boomers. To do so, we averaged data from each available age stratum that encompassed those born during the baby boom, weighted by the number of individuals in that stratum, and compared those data to prevalence rates among individuals under age 35.
In 1979 overall 27 percent of younger baby boomers, age 21 to 33—almost 14 million people—reported using any illicit drug during the past month. As baby boomers aged, the prevalence of the use of any illicit drug in the past month declined sharply until the individuals reached their early thirties, when it leveled out. The prevalence rate has remained stable but higher than age-matched cohorts from previous generations. These data suggest that some baby boomers have continued to use illicit drugs as they age. Due to their sheer numbers, we can expect larger numbers of current drug users to reach age 65 and to have a potential impact on treatment programs and other resources.
Alcohol abuse
Liberto and associates (
36) reviewed literature on rates of alcohol abuse and dependence among people over age 65 and estimated that the incidence of heavy drinking (12 to 21 drinks a week) was between 3 percent and 9 percent. However, the ECA study, which used
DSM-III criteria, estimated a much lower one-month prevalence rate of alcohol abuse and dependence—.9 percent—among people over age 65. Another study of urban residents also found a low one-month rate of 2.2 percent in this age group (
37). The National Longitudinal Alcohol Epidemiologic Survey estimated the prevalence of alcohol dependence among persons over age 65 to be 1.2 percent for men and .3 percent for women (
38).
Hospitalized individuals have rates of incidence of alcohol abuse that are higher than those in community samples, with estimates ranging from 5 percent to 50 percent (
36). Some longitudinal studies suggest that alcohol consumption decreases with age (
21,
39), while others have reported stable consumption (
40) or increased consumption (
41). According to Reid and Anderson (
42), alcohol abuse and dependence are increasingly being recognized as problems among elderly persons. Prevalence rates are likely to increase as baby boomers, who have heavier drinking habits than the current cohort of older adults, reach older age.
Age-related changes in substance abuse
Several factors may be related to the observed age-related reductions in substance abuse. They include problems with diagnostic criteria associated with aging, age-related changes in pharmacokinetics, and changing patterns of drug use—for example, reductions in illicit drug use. Winick (
43) proposed one of the most popular theories— "maturing out"—to explain apparent decreases in substance abuse, particularly of narcotics, associated with aging. This theory posits that factors associated with aging processes and length of abuse contribute to a decline in the number of older narcotic addicts. These factors include age-related developmental changes and morbidity and mortality associated with substance use.
Substance abusers have higher mortality rates than age-matched nonabusers (
23,
44). Studies of older individuals have reported that light and heavy drinkers are at an increased risk of mortality, while moderate drinkers are at a decreased risk (
45,
46). According to the maturing-out theory, increased numbers of individuals addicted to illicit drugs become abstinent as they age. However, this hypothesis is not supported by empirical data, which suggest that the number of older addicts may increase and that persons who have been addicted for more than five years do not become abstinent as they age (
32,
47). One study reported that addicts approaching age 50 who were followed for more than 20 years remained involved in criminal activities (
32).
A number of other factors have been associated with the risk of substance abuse among elderly persons. They include biologic factors, such as changes in drug metabolism noted above. Other researchers have pointed to demographic characteristics such as gender as important risk factors. However, the relationship between gender and risk of substance abuse does not appear to be simple. Although women are more likely to receive a prescription for psychoactive drugs (
22,
48), men may use psychoactive drugs more frequently than women after age 65 (
49).
Several psychosocial variables have been hypothesized to be risk factors for substance abuse in the elderly population. Stress (
50), isolation (
51,
52), various losses, loneliness, and onset of illness have all been related to late-onset drug use (
52). Finlayson (
53) noted that for older persons who abuse alcohol the roles of social support, social control, and coping might be different than for those who abuse other drugs. For example, an individual with a physical problem such as pain may be encouraged by friends and family to use pain medications, while individuals with alcohol dependence may find that support persons reject their use of alcohol.
In addition to physical comorbidity, use of psychotropic drugs and psychiatric comorbidity are important risk factors for substance abuse. As many as one-fourth of elderly persons living in the community use psychotropic drugs (
15,
16,
53). Psychiatric diagnoses that have been associated with prescription drug dependence include a personality disorder, somatoform disorders, and anxiety, sleep, and adjustment disorders (
17).