Skip To Content
Click for DHHS Home Page
Click for the SAMHSA Home Page
Click for the OAS Drug Abuse Statistics Home Page
Click for What's New
Click for Recent Reports and HighlightsClick for Information by Topic Click for OAS Data Systems and more Pubs Click for Data on Specific Drugs of Use Click for Short Reports and Facts Click for Frequently Asked Questions Click for Publications Click to send OAS Comments, Questions and Requests Click for OAS Home Page Click for Substance Abuse and Mental Health Services Administration Home Page Click to Search Our Site

Substance Abuse by Older Adults:  Estimates of Future Impact on the Treatment System

Table Of Contents

Chapter 3. Drug Use Careers: Recovery and Mortality

Yih-Ing Hser,* Ph.D.

Abstract: Although long-term follow-up studies of substance-using individuals provide direct measurement of recovery and mortality, few in the literature report age-specific recovery and mortality rates. This chapter examines these parameters based on data from two studies. The California 33-year Follow-up Study of 581 heroin addicts showed a 49 percent overall mortality rate, with rates increasing with age from about 33 percent for those in the 45- to 49-year age category to 76 percent or higher among those older than 65. Rates of permanent recovery (i.e., abstinent for 5 years or more) also appeared to increase from 36 percent for those in the 45- to 49-year age category to about 50 percent for older ages, but there was no continuing increase of recovery after age 50. The Study of High-risk Samples (i.e., patients from emergency rooms and sexually transmitted disease clinics, recent arrestees) showed that baby boomers generally reported greater levels of illicit drug use but comparable levels of use of medications. The relationship between age and the drug use progression or recovery process appears to differ depending on the type of drug used. Future studies need to include women and to improve sampling of the elderly.

 

Introduction

The purpose of this chapter is to provide direct measurement of recovery and mortality in the drug use careers of two special populations: (a) male opiate users and (b) drug users identified at high-risk sites (emergency rooms [ERs], sexually transmitted disease [STD] clinics, and jails). The male opiate users have been followed longitudinally in a California 33-year Follow-up Study (Hser, Hoffman, Grella, & Anglin, 2001). The samples in the newer Study of High-risk Samples include users of various types of drugs, but have been followed for only 3 years so far (Hser, Boyle, & Anglin, 1998). These data provide the basis for an understanding of the longitudinal changes in drug use as drug users age, with a particular focus on mortality and recovery rates and the baby boom generation (i.e., those born between 1946 and 1964).

Direct population-based measures of drug use recovery rates and mortality rates are unavailable. This chapter summarizes estimates from the two previously mentioned studies and from those published based on long-term follow-up studies of drug users from both treated and nontreated samples. Each of these studies was designed to address specific issues, and their samples are often considered special populations that may be limited in sample size and representation (demographics, geographical location, etc.). Nevertheless, the studies provide longitudinal estimates of recovery and mortality rates associated with drug use, which are critically needed for projecting future health care needs among populations in need as they age.

This chapter first reviews existing longitudinal studies that provide estimates of recovery and mortality rates among substance abusers. Then recovery and mortality rates are reported based on a 33-year follow-up of heroin users, as well as rates of drug use and changes (e.g., relapse and quitting) over three annual assessment points among a high-risk sample of drug users recruited from nontreatment sources.

 

Background: Mortality and Recovery Rates

Several long-term follow-up studies of substance abusers have been conducted in the United States. For the specific purpose of this chapter, the most representative studies published in the past 5 to 6 years are reviewed (i.e., those with relatively large sample sizes and follow-up periods of 3 years or longer). Studies that are follow-ups of treatment samples are differentiated from those that are not because recovery and mortality rates may be influenced by the treatment intervention. Sample characteristics and findings on recovery and mortality rates of these studies are summarized in Table  1.

 

Treatment Sample

SROS 5-year follow-up of a national treatment sample. The Services Research Outcomes Study (Office of Applied Studies [OAS], 1998) reported a mortality rate of 9.1 percent (277 deaths among a targeted sample of 3,047) over 5 years posttreatment. The study was based on a national probability sample of 99 treatment programs and 3,047 patients discharged from these programs in 1989 and 1990. Treatment program modalities were hospital inpatient, residential, outpatient methadone (including detoxification and maintenance), and outpatient nonmethadone programs. The study conducted record abstraction of these patients during 1994 and follow-up interviews during 1995-96. The ages of the patients ranged from 15 to 60 years at treatment discharge in 1991. Based on a fairly small number of deaths, 277, when compared with the number of deaths in the total population over 5 years, these substance abusers had estimated mortality rates that were 7.3 times the general population of similar age, gender, and race; the death rate was 8 times for white males, 18 times of that for white females, 5 times for black males, and 7 times for black females.

 

 

Table 1 Sample Characteristics and Mortality and Recovery Rates of Studies of Alcohol- and Drug-Abusing Populations

Study Name

Geographic
Coverage

Sample Size
(Total/Follow-Up)1

Age Range2

Gender
(Male/Female)

Ethnicity
(White/Black/Other)

Years of Follow-Up

Recovery3

Mortality

Treatment Sample

               
     SROS

National

3,047 / 1,799

15-60

71 / 29

60 / 28 / 12

5

21%

9%

     DATOS

National

2,147 / 1,042

21-74

63 / 37

42 / 46 / 12

5

18%

6%

     Moos' VA Study

National

21,139 / N/A

55-75+

99 / 1

83 / 13 / 4

4

N/A

24%

Nontreatment Sample

               
     Vaillant's Study

Local

55 / 52

47

100 / 0

100 / 0 / 0

13

11%

18%

   

150 / 129

47

100 / 0

100 / 0 / 0

13

30%

28%

     California 33-
     year Follow-up
     Study4

California

581 / 242

18-45

100 / 0

37 / 7 / 56

33

47%

49%

N/A: Not applicable; DATOS: Drug Abuse Treatment Outcome Studies; SROS: Service Research Outcome Study.

1 Total sample size was the denominator for the mortality rate, and the follow-up sample size was the denominator for calculating the recovery rate.
2 Ages refer to those at the baseline of the follow-up study.
3 "Recovery" is defined as abstinent for 5 years or more at the time of final assessment and was based on the follow-up sample; one exception is Vaillant's study, which reported 3-year abstinence rates for those aged 60 or older.
4 Although the original sample was recruited from a compulsory treatment program for criminal offenders, we consider this a nontreatment sample because data reported in this chapter are mostly based on the long-term follow-up that was conducted more than 33 years post the original program admission.

Self-reported data on drug use from 1,799 patients who completed the follow-up interview indicated that those using any illicit drug decreased from 75 percent before treatment to 59 percent 5 years after treatment, which was a statistically significant difference of minus 16 points, or a 21 percent recovery rate.

DATOS 5-year follow-up of a community treatment sample. The Drug Abuse Treatment Outcome Studies (DATOS) reported a 6.0 percent mortality rate (based on 128 deaths among the 2,147 targeted sample) over 5 years posttreatment (Hubbard, Craddock, & Anderson, 2001). The original sample was recruited from 1991 to 1993 at DATOS treatment admission to 96 programs, which involved short-term inpatient, long-term residential, methadone maintenance, and outpatient drug-free programs. The mean age at treatment admission was 32.6 years (standard deviation [SD] = 7.6); at the 5-year follow-up interview, the mean age was 40.0 (SD = 7.7).

Patterns of abstinence and relapse (N = 1,042) at the 5-year follow-up were as follows: Approximately 18 percent had been abstinent without relapse, 14 percent relapsed once and then stopped use, 27 percent relapsed and never stopped using, and 41 percent relapsed, stopped use, and relapsed again. About 32 percent reentered treatment after DATOS discharge (Hser, Grella, Shen, & Anglin, 2000). Additionally, at the 5-year follow-up interview, 66.4 percent were not using marijuana, 64.5 percent were not using cocaine, and 77.6 percent were not using heroin (compared with the respective rates of 47.5, 30.6, and 59.2 percent during the year before treatment).

Moos et al.'s VA study. A mortality rate of 24 percent (2.64 times higher than expected) was reported among late-middle-aged and older (55+) substance abuse inpatients (N = 21,139) in Department of Veterans Affairs (VA) Medical Centers who were followed for 4 years after receiving inpatient care (Moos, Brennan, & Mertens, 1994). The study was based on VA records of patients who were diagnosed with substance abuse disorder and were discharged from VA inpatient programs during a 1-year period (October 1, 1986, to September 30, 1987). Age-specific mortality rates were as follows: 21.3 percent for those aged 55 to 64 years, 29.4 percent for those aged 65 to 74 years, and 45.4 percent for those aged 75 or older.

 

Nontreatment Sample

Vaillant's follow-up study of alcohol abusers. A prospective study of the alcohol use of two samples of men (268 college students, 456 city adolescents) has been conducted since 1940 and followed up for 50 years (Vaillant, 1996). By age 70 years, 52 (21 percent) of the 249 college men who remained in the study at age 47 years had met the DSM-III criteria for alcohol abuse (as had 3 of the 13 who dropped out of the study); at some point, 21 of the 55 men (38 percent) also met the criteria for dependence.1 By age 60 years, 150 of the 433 city men (35 percent) who remained in the study at age 47 years and whose alcohol abuse status was known had met the DSM-III criteria for alcohol abuse. At some point, 77 (51 percent) of the city alcohol abusers also met the criteria for alcohol dependence. The average age at onset of alcohol abuse for the 51 college men was 40.2 (SD=9.9) years; for the city men, it was 29.2 (SD=9.5) years.

By 60 years of age, 18 percent of the college alcohol abusers had died, 11 percent were abstinent, 11 percent were controlled drinkers, and 59 percent were known to be still abusing alcohol. By 60 years of age, 28 percent of the city alcohol abusers had died, 30 percent were abstinent, 11 percent were controlled drinkers, and only 28 percent were known to be still abusing alcohol. Vaillant (1996) concluded that after abstinence had been maintained for 5 years, relapse was rare. In contrast, return to controlled drinking without eventual relapse was unlikely.

California 33-year Follow-up Study. The sample of Californian heroin users who were followed for 33 years consisted of 581 male narcotics/opiate addicts admitted to the California Civil Addict Program (CAP) from 1962 through 1964. Although the original sample was recruited from CAP, the sample is considered a nontreatment sample because data reported in this chapter are mostly based on the long-term follow-up conducted more than 33 years after the original program admission. Mean age at admission was 25.4 years. Three face-to-face interviews were conducted with the sample every 10 years over 33 years (Hser et al., 2001).

At the first follow-up study in 1974-75, 13.8 percent of the original 581 sample had died. The average age of the 439 living addicts interviewed at that time was 36.8 years. Urine tests revealed that 37.8 percent of the interviewed sample showed no opiate use, although they could have been using other drugs. At the second follow-up study in 1985-86, 27.7 percent of the original 581 sample had died, and the average age of the 354 interviewed was 47.6. Urinalysis revealed that 41.0 percent of those interviewed tested negative for opiates. At the latest, or third follow-up study in 1996-97, close to half (48.9 percent) of the original sample had died. The average age of the 242 who were interviewed was 57.4, and 55.8 percent of them tested negative for opiates.

Among the 284 confirmed deaths, the most common cause of death (21.6 percent) was accidental poisoning (ICD-9 code 850.0)2 or drug overdose. A total of 45 subjects' death certificates specified overdose due to heroin, and 16 were due to use of other drugs. The next most common causes of death were chronic liver disease (15.2 percent), cancer (11.7 percent), and cardiovascular diseases (11.7 percent). Other deaths included homicide (8.2 percent), accidents (8.5 percent), or suicide (2.8 percent).

Among the 242 subjects interviewed in 1996-97, 46.7 percent reported continuous abstinence from heroin use in the past 5 or more years. Their long-term heroin abstinence was associated with less criminality, morbidity, psychological distress, and higher employment.

Although the literature is limited, the brief review presented here provides a broad picture of mortality and recovery rates among alcohol abusers or illicit drug users. However, few of these studies provide these parameters by specific age categories, which will be needed if projection into the future is desired. The following sections examine age-specific mortality and recovery rates in two studies—the California 33-year Follow-up Study of a sample of heroin users and the Study of High-risk Samples.

 

California 33-year Follow-up Study

 

Study Design

As noted earlier, the California 33-year Follow-up Study was a prospective longitudinal study of 581 male narcotics addicts admitted to CAP from 1962 through 1964. Established in 1961 by California legislation, CAP was a compulsory drug treatment program for narcotics-dependent criminal offenders committed under court order. The program consisted of an inpatient period followed by supervised community aftercare. Patients could be returned for further inpatient stays if there was evidence of relapse to addiction or other behaviors that violated conditions of aftercare. This program offered the only major publicly funded treatment to California addicts during the 1960s; during the 1970s, methadone maintenance became commonly available. The sample, selected from the 1962 through 1964 admission records, was first interviewed in 1974-75 as part of an evaluation of CAP (McGlothlin, Anglin, & Wilson, 1977). A second follow-up of this sample was conducted in 1985-86 (Hser, Anglin, & Powers, 1993), and a third in 1996-97 (Hser et al., 2001).

 

Sample

The sample of 581 heroin users was limited to male subjects because of the small number of female commitments to CAP. The sample consisted of white (36.5 percent), Hispanic (55.6 percent), and African-American (7.9 percent) addicts. Before age 18, more than 80 percent of the sample had been arrested, and 80 percent had tried marijuana. More than 60 percent of the sample started using narcotics before age 20. Mean age at admission in 1962-64 was 25.4 years (SD=3.9). The mean age of the 242 subjects interviewed in 1996-97 was 57.4 years (SD=4.0). The 1996-97 study had a 96 percent location rate (242 interviewed, 31 refused or were too mentally dysfunctional to be interviewed, and 284 were confirmed to be dead), with 24 subjects lost to follow-up.

 

Interview Procedure

The three face-to-face interviews conducted at 10-year intervals collected information on patterns of drug use and related activities. The interview protocol was adapted from Nurco, Bonito, Lerner, and Balter (1975) and was designed to obtain information on subjects' demographic characteristics, family history, drug use history, employment, and criminal behavior, as well as information on their legal status history (incarcerated, under legal supervision but not incarcerated, and unsupervised). Subjects were aware that the interviewer already knew their official history of criminal activity and legal status from information obtained independently from California criminal justice system records and could therefore verify the subjects' self-reports of criminal activity and legal status. Subjects gave informed consent for study participation and were given written assurances of confidentiality prior to the interview. Multiple measures were retrospectively recalled in a chronological sequence that covered, across three interviews, from 1 year prior to their first narcotics use to the time of the 1996-97 interview. The average interview at each follow-up point required between 2 and 3 hours to administer. At the end of each interview, a urine specimen was collected from those subjects who were not incarcerated. All participation, including the furnishing of urine samples, was voluntary.

The rates of congruence between self-reported current opiate use and urinalysis among those who provided a urine specimen was 73.7 percent at the first interview, 85.8 percent at the second interview, and 90.2 percent at the third interview. The reliability of the instrument has been examined and reported elsewhere (Anglin, Hser, & Chou, 1993; Hser, Anglin, & Chou, 1992).

 

Results

Rates of mortality and recovery (i.e., abstinent from heroin for 5 years or more) were calculated by 5-year age categories constructed from the age of the subjects (alive or dead) in 1992 (see Figure 1). This year was selected so as to enable comparisons across the chapters included in this monograph. As expected, rates of death increased steadily as a function of the ages of the respondents (see Table  2). Permanent recovery also appeared to increase from 36.5 percent at the age category of 45 to 49 years to about 50 percent at older ages, but there did not appear to be a continuing increase of recovery after age 50. (Unfortunately, partly because of the high mortality rates, the numbers of subjects at the age categories of 60 years or older were too small to support reliable estimates of recovery.)

 

 

Figure 1 California 33-year Follow-up Study Sample Distribution, by Age, in 1992

 D

 

 

Table 2 Mortality and Recovery Rates, by Age Group, in the California 33-year Follow-up Study Sample

Age Category in 1992

Number of Subjects (N = 581)

Deaths by 1997 (%)

Number of Subjects Interviewed in 1997 (N = 242)

Abstinent for 5 Years
or More (%)

45 to 49

117

33.3

63

36.5

50 to 54

245

41.2

117

49.6

55 to 59

124

50.0

47

53.2

60 to 64

63

74.6

12

41.7

65 to 69

21

76.2

2

50.0

70 to 74

5

80.0

1

100.0

75 or Older

6

100.0

0

--

 

Study of High-risk Samples

 

Study Design

The Study of High-risk Samples was also a prospective longitudinal study, similar to the California 33-year Follow-up Study, but it examined other drug use as well as opiate use among an expanded range of high-risk populations (Hser et al., 1998; Hser, Maglione, & Boyle, 1999). In 1992-94, more than 5,000 individuals were interviewed and screened in ERs, STD clinics, and jails in Los Angeles County (Hser et al., 1998). Annual follow-up interviews were conducted with successively smaller (randomly selected) subsets of the original sample (i.e., approximately 900 drug users at Follow-up 1, 500 at Follow-up 2, and 300 at Follow-up 3). The criterion for inclusion of young adults aged 18 to 25 years in the follow-up was use of any illegal drug in the past year or a history of ever being dependent on an illegal drug. For subjects 26 years or older, the criterion was past year use of an illegal drug other than marijuana or a history of ever being dependent on an illegal drug. These selection criteria allowed oversampling of young adults, many of whom were still in an early stage of a drug use career.

 

Subjects

The study examined 5,168 individuals at baseline (1,571 patients in ERs, 1,563 patients in STD clinics, and 2,034 arrestees in jails). Subjects were screened and assessed through several stages. The intake sites included three ERs, three STD clinics, and four county jails, all located in Los Angeles County. Study sites were purposefully selected, as opposed to randomly selected, to provide a sufficient subject pool and to include gender and ethnic diversity.

Interviewers visited ERs approximately 2 days a week on various days and at different times of day and night in an effort to capture the full range of patients. Subjects were randomly selected from the sign-in list. The refusal rate was 14.6 percent of all subjects approached. A total of 1,571 patients from ERs were interviewed (39.2 percent female, 14.3 percent white, 40.4 percent Hispanic, 40.4 percent African American, and 5.0 percent other races/ethnicities, with a mean age of 37.0 years).

In STD clinics, research interviewers used the sign-in list as a base and attempted to interview every person on the list. Overall, about 23 percent of those approached refused to participate in the study. A total of 1,563 patients were interviewed (40.6 percent female, 7.1 percent white, 40.1 percent Hispanic, 50.5 percent African American, and 2.3 percent other races/ethnicities, with a mean age of 29.7 years).

The four jails were sites included in the National Institute of Justice (NIJ) Drug Use Forecasting (DUF) study. Interviewers collected data for 1 week at each jail on a quarterly cycle. The subject selection was in accordance with the DUF protocol, which had the following prioritization for subject selection: those arrested for nondrug felonies, followed by nondrug misdemeanors, then drug felonies, and finally drug misdemeanors. About 6 percent of those approached refused to participate. A total of 2,034 adult arrestees were interviewed (34.1 percent female, 19.6 percent white, 41.7 percent Hispanic, 35.7 percent African American, and 2.7 percent other races/ethnicities, with a mean age of 30.1 years). Age distribution in 1992 for the total sample is displayed in Figure 2.

 

 

Figure 2 Study of High-risk Samples Distribution, by Age, in 1992

 D

 

Data Collection Procedures

Face-to-face interview procedures and questionnaires were similar across sites. At STD and ER sites, study participants were paid $5 for their participation. Interviews with arrestees in jail went uncompensated, but snacks were offered as an inducement. The baseline interview lasted approximately 25 minutes, at the end of which a urine specimen was collected (including those in jails). Each subsequent face-to-face interview lasted 2 to 2½ hours, and respondents were paid $40 to $50 for each completed interview. A urine specimen was obtained from those not incarcerated at the end of these interviews. Respondents were queried about demographics, personality measures, drug use history, and other measures. Respondents were assured that all information provided to the researchers would be held in the strictest confidence.

 

Results

Prevalence rates among the high-risk samples. Prevalence rates of self-reported lifetime alcohol, tobacco, and other drug use are provided by 5-year age categories constructed from the age of the participants in 1992 (Table  3). Not surprisingly, almost all respondents reported alcohol use, and to a lesser extent, tobacco use sometime in their lifetime. In terms of illicit drugs, marijuana and crack/cocaine were the most prevalent drugs used by all age groups. In general, compared with other age categories, the baby boomers (approximately 28 to 46 years of age) reported a higher level use of all drugs: marijuana and cocaine use at rates between 45 and 75 percent, followed by heroin, PCP, amphetamines, Valium, and downers at about 20 to 30 percent. Older adults aged 65 or older reported virtually no illicit drug use.

Self-reported use in the past 30 days (Table  4) followed similar patterns, although at much lower rates. Table  4 shows that in this high-risk population, use of such drugs as speed, PCP, amphetamines, barbiturates, tranquilizers, and LSD stopped by age 55. Marijuana, heroin, and crack and cocaine use stopped before age 65. Urinalysis results for recent use confirmed that crack/cocaine was the most frequently used illicit drug by this sample, with the highest rates among baby boomers (33 to 42 percent) (Table  5).

A subgroup of the sample (1,313 patients from ER sites, 767 from STD sites, and 142 from jail sites) also responded to a questionnaire on their use of 11 types of medications (Table  6). Rates of self-reported use in the past 30 days were generally low, except for pain pills (nonopiate), a use of which was reported by at least 25 percent of the respondents across all age groups. The use patterns among baby boomers appeared to be comparable with other age groups, including the elderly.

 

 

Table 3 Self-Reported Lifetime Use (Percent) at Baseline (1993-94) in the Study of High-risk Samples (N = 5,155)

Age Category (in 1992)

N

Alcohol

Tobacco

Marijuana

Heroin

Crack

Cocaine

PCP

Speed

Amphet-
amines

Barbi-
turates

Tran-
quilizers

LSD

15-19

438

94.75

71.00

68.04

1.60

18.72

24.20

14.61

12.56

4.79

2.51

3.88

15.75

20-24

1,165

95.36

71.07

65.32

4.29

21.63

28.33

12.79

12.02

7.98

5.24

8.93

13.30

25-29

1,004

95.82

79.78

74.20

11.95

40.04

46.81

23.21

18.43

15.84

11.35

17.83

18.73

30-34

932

95.28

86.37

75.75

14.91

49.25

53.86

31.97

18.13

17.17

15.34

26.61

19.53

35-39

668

97.75

85.03

74.25

22.46

49.85

55.24

32.78

18.26

28.44

28.29

35.48

25.15

40-44

389

95.12

84.58

67.10

23.65

46.27

45.76

24.42

14.65

25.71

28.02

32.90

26.48

45-49

218

95.41

84.40

61.47

25.69

33.49

41.74

19.72

14.22

26.15

26.61

31.65

22.48

50-54

177

93.79

79.66

49.15

15.82

22.03

25.99

7.91

10.17

14.69

15.25

20.90

9.60

55-59

74

89.19

79.73

25.68

5.41

8.11

13.51

2.70

2.70

10.81

6.76

13.51

4.05

60-64

48

75.00

70.83

20.83

6.25

8.33

6.25

4.17

2.08

4.17

8.33

8.33

4.17

65-69

20

90.00

80.00

25.00

5.00

10.00

5.00

0.00

0.00

5.00

5.00

10.00

0.00

70-74

11

90.91

72.73

18.18

0.00

0.00

9.09

0.00

0.00

0.00

0.00

0.00

0.00

75-79

8

62.50

62.50

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

12.50

0.00

80-84

2

50.00

50.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

85-89

1

100.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

 

 

Table 4 Self-Reported Drug Use in the Past 30 Days (Percent) at Baseline (1993-94) in the Study of High-risk Samples (N = 5,155)

Age Category (in 1992)

N

Alcohol

Tobacco

Marijuana

Heroin

Crack

Cocaine

PCP

Speed

Amphet-
amines

Barbi-
turates

Tran-
quilizers

LSD

15-19

438

63.93

45.89

38.58

0.46

8.45

6.62

3.88

4.57

0.46

1.82

0.68

1.83

20-24

1,165

69.36

46.87

34.76

1.37

10.30

6.09

2.75

4.29

1.29

1.12

1.03

1.12

25-29

1,004

69.02

57.27

32.17

3.29

20.62

10.46

1.79

5.38

2.09

1.29

2.29

0.40

30-34

932

64.27

64.48

25.21

4.29

25.64

10.19

2.15

4.29

1.29

0.75

2.04

0.43

35-39

668

62.72

62.28

21.26

4.04

26.50

9.58

2.54

3.74

1.50

2.10

4.64

0.00

40-44

389

62.21

63.24

20.82

5.14

28.53

10.03

2.57

1.03

1.03

1.03

3.34

0.00

45-49

218

55.05

54.59

13.76

5.05

16.06

5.96

0.00

1.83

1.38

0.92

3.21

0.00

50-54

177

48.02

48.02

9.04

3.39

8.47

3.95

0.00

2.26

0.56

0.56

2.26

0.00

55-59

74

45.95

33.78

2.70

0.00

2.70

2.70

0.00

0.00

0.00

0.00

0.00

0.00

60-64

48

22.92

31.25

6.25

2.08

4.17

2.08

0.00

0.00

0.00

0.00

0.00

0.00

65-69

20

35.00

5.00

0.00

0.00

5.00

0.00

0.00

0.00

0.00

0.00

5.00

0.00

70-74

11

27.27

9.90

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

75-79

8

25.00

37.50

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

80-84

2

50.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

85-89

1

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

 

 

Table 5 Urinalysis Positive Rates (Percent) at Baseline (1993-94) in the Study of High-risk Samples (N = 4,503)

Age Category (in 1992)

N

Benzo-
diazepines

Marijuana

PCP

Opiates

Cocaine

Methadone

Barbiturates

Amphet-
amines

15-19

379

1.32

21.64

3.96

2.90

12.66

0.00

0.26

2.64

20-24

1,036

1.64

22.10

3.38

3.19

18.73

0.19

0.10

3.96

25-29

892

2.35

18.27

2.35

6.05

33.30

0.45

0.45

4.60

30-34

842

4.04

15.20

2.73

8.91

42.04

0.95

0.00

5.23

35-39

596

6.38

11.24

3.02

9.40

41.11

0.50

0.34

3.52

40-44

334

4.19

10.78

2.10

11.38

38.92

2.10

3.89

1.50

45-49

181

9.39

7.73

0.55

13.81

29.28

2.21

0.55

4.42

50-54

135

6.67

8.15

1.48

9.63

20.00

3.70

2.22

1.48

55-59

56

3.57

3.57

1.79

7.14

10.71

0.00

3.57

0.00

60-64

35

5.71

2.86

0.00

5.71

11.43

2.86

0.00

0.00

65-69

10

0.00

0.00

0.00

10.00

10.00

0.00

0.00

0.00

70-74

3

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

75-79

4

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

 

 

Table 6 Self-Reported Use of Medications (Percent) in the Past 30 Days at Baseline (1993-94) in the Study of High-risk Samples (N = 2,222)

Age Group

N

Pain Pill—
Nonopiate

Pain Pill—
Opiate

Sleeping Pill

Tran-
quilizer

Anti-
depressant

Diet, Pick-up—
OTC

Diet, Pick-up, Prescription

Cold,
Allergy—
Nonopiate

Cold,
Allergy—
Opiate

Seizure Medi-
cations

Blood Pressure Medication

15-19

198

32.32

5.05

1.01

1.52

0.00

0.51

0.51

7.58

0.51

0.51

0.00

20-24

442

27.38

5.43

1.13

0.45

0.23

1.13

0.68

6.33

0.45

0.90

0.23

25-29

373

30.03

5.36

1.88

1.34

1.61

1.61

0.27

7.51

0.54

1.07

1.07

30-34

367

26.98

9.26

0.54

1.91

0.82

0.00

0.00

4.36

0.27

0.82

1.91

35-39

303

27.06

6.93

0.33

1.98

1.32

0.33

0.33

4.62

0.00

1.32

2.64

40-44

196

28.57

7.14

2.04

4.08

3.57

0.00

0.00

8.16

0.51

4.08

4.59

45-49

117

29.06

8.55

0.85

3.42

1.71

0.00

0.00

4.27

0.85

4.27

9.40

50-54

108

31.48

8.33

0.93

2.78

2.78

0.00

0.00

6.48

0.93

0.00

11.11

55-59

45

31.11

2.22

2.22

2.22

0.00

0.00

0.00

0.00

0.00

0.00

17.78

60-64

36

44.44

0.00

2.78

2.78

0.00

0.00

0.00

5.56

0.00

0.00

27.78

65-69

18

38.89

0.00

0.00

11.11

5.56

0.00

0.00

0.00

0.00

0.00

5.56

70-74

8

25.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

12.50

75-79

8

25.00

0.00

0.00

0.00

0.00

0.00

0.00

12.50

0.00

12.50

12.50

80-84

2

50.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

50.00

85-89

1

100.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

Drug use careers. Many of these high-risk respondents used various types of drugs on a regular basis. Using the Follow-up 1 sample (N = 941) with drug users from three sources, approximately 75 percent of them used alcohol and marijuana regularly, followed by crack at 50 percent, cocaine at 34 percent, and opiates at about 17 percent. Natural histories of drug use were collected during Follow-up 2 (N = 564), and the rates of any weekly use of cocaine, heroin, and marijuana are shown over the evolution of the drug use careers (Figure 3). The rates of marijuana use showed a clearly linear decline as the cohort aged. Rates of cocaine use increased from age 20 until the mid-30s, then declined after the late 30s. Heroin use, on the other hand, remained at a fairly low level, but increased after the early 40s. A further examination indicates that fewer than 50 subjects were older than 45 years old, and most of these older respondents were heroin users; thus, they appeared to have contributed to a greater proportion of the later segment of the drug use careers and disproportionally increased the heroin use rates as the group aged.

 

 

Figure 3 Self-Reported Any Weekly Use of Illicit Drugs, by Age, in 1992

 D

Relapse and remission among drug users. Because the sample has only been followed for 3 years, permanent recovery rates (i.e., 5-year abstinence) cannot be established for this sample. However, patterns of drug use over time can be studied year by year and over the three observation points. The probabilities of changing drug use status, given the drug use status in the previous year, are given in Table  7. Drug use included any use of illicit drugs in the year before an interview. Because each successive follow-up sample had a smaller sample size, the probabilities were calculated repeatedly for each of the three samples with varying observation points for replication purpose. The three samples did not appear to differ much in rates where comparable measures were available; thus, the Follow-up 3 sample was focused on because transition probabilities were available over 3 follow-up years. The likelihood of continued drug use over 2 consecutive years was very high, ranging from .59 to .70. The probabilities of quitting (i.e., stopping drug use during the current year among those who reported drug use in the previous year) were .41 in Follow-up 1, .30 in Follow-up 2, and .30 in Follow-up 3. High relapse rates (i.e., drug use during the current year while reporting no drug use during the previous year) were observed at .40 at Follow-up 2 and .27 at Follow-up 3.

 

 

Table 7 Yearly Changes in Illicit Drug Use Status among Drug Users (Conditional Probabilities)

 

Follow-up 1 Sample

Follow-up 2 Sample

Follow-up 3 Sample

Baseline to Follow-up 1

     

    Persist using

.55

.57

.59

    Quit

.45

.42

.41

    Relapse

N/A

N/A

N/A

Follow-up 1 to Follow-up 2

     

    Persist using

 

.72

.70

    Quit

 

.28

.30

    Relapse

 

.37

.40

Follow-up 2 to Follow-up 3

     

    Persist using

   

.70

    Quit

   

.30

    Relapse

   

.27

 

Comments

Long-term follow-up studies provide direct measurement of recovery and mortality in drug use careers. The California 33-year Follow-up Study is unique in that age-specific rates of recovery and mortality can be calculated for heroin users. The newer Study of High-risk Samples expands the longitudinal examination of drug use beyond heroin and can begin to provide some preliminary estimates of drug use relapse and remission. However, the two studies are not without limitations. The California 33-year Follow-up Study did not include women, and most of these heroin users were born before the period of the baby boomers. The Study of High-risk Samples had short follow-up periods, the sample decreased at each follow-up point, and the study did not include an adequate sample of older drug users. Samples included in both studies should not be considered nationally representative.

Nonetheless, both studies reveal several interesting results. The Study of High-risk Samples suggests that baby boomers generally showed greater levels of illicit drug use but comparable levels of use of medications. Another important finding is that the drug use progression and recovery process appears to differ depending on the type of drug used. Heroin use appeared to be most persistent with rates of permanent recovery remaining at about 50 percent even among those older than 50, with rates of use even increasing after age 40 among the high-risk samples. Furthermore, among the high-risk samples, marijuana use was linearly related to age (i.e., decreasing with age), and cocaine use and age were curvelinearly related (i.e., peaking at about ages 35 to 40). Had drug type been ignored, the group would have been extremely stable in drug use across ages.

The ranges of recovery and mortality rates appeared to be wide across these two studies and those reported in the literature. It is also difficult to compare these studies, as each study had different sample compositions (e.g., age, gender, alcohol or other drug type) and methodologies (e.g., length of follow-up period, retrospective vs. prospective study design). The morality rates ranged from 6 and 9 percent in DATOS and SROS, to 24 percent in the Moos VA study, and to 49 percent in the California heroin sample. Both the Moos study and the California study focused on older adults; therefore, it is not unexpected their death rates were higher than the treatment samples in DATOS or SROS where most of the patients were in their 30s or 40s. The relatively higher mortality rate in the California study than that in the Moos study, even after adjusting for age (Table  1) suggests that heroin users are at an even greater risk for premature death. However, it should be noted that the sample sizes for calculating mortality rates were small in most of these studies when compared with population estimates.

The recovery rates also showed a diverse range of estimates. Vaillant's alcohol abuse study reported 11 and 30 percent for recovery rates, and the California heroin sample showed 47 percent, while more general treatment samples in DATOS and SROS suggested 18 and 21 percent recovery rates. Even using the same database, different definitions of recovery also contribute to discrepant estimates. For example, in another chapter in this monograph, Ray used treatment readmission to operationally define "recovery" and reported a 50 percent recovery rate among patients followed in the SROS, while the present chapter defined recovery as "abstinence from any illicit drug use," which suggested a 21 percent recovery rate for SROS patients. Using yearly change rates across 3 follow-up years, the Study of High-risk Samples showed an overall yearly quitting rate at about 30 to 40 percent.

One obvious conclusion from these diverse findings is that data on age-specific recovery and mortality rates are needed to permit standardized comparisons, but these data are scarce at the present time. Much more research is needed to fill in these gaps so that informed decisions can be made on projecting future health care needs among populations as they age. Particularly, future studies addressing recovery and mortality issues need to improve sample representativeness, especially the sampling of women and the elderly, and they need to pay attention to the different types of drugs used by the individuals.

 

Conclusions

 

References

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd rev. ed.). Washington, DC: Author.

Anglin, M. D., Hser, Y., & Chou, C.-P. (1993). Reliability and validity of retrospective self-report by narcotics addicts. Evaluation Review, 17(1), 90-108.

Hser, Y. I., Anglin, D., & Powers, K. (1993). A 24-year follow-up of California narcotics addicts. Archives of General Psychiatry, 50, 577-584.

Hser, Y. I., Anglin, M. D., & Chou, C.-P. (1992). Reliability of retrospective self-report by heroin addicts. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 4, 207-213.

Hser, Y. I., Boyle, K., & Anglin, M. D. (1998). Drug use and correlates among sexually transmitted disease patients, emergency room patients, and arrestees. Journal of Drug Issues, 28, 437-454.

Hser, Y. I., Grella, C., Shen, H., & Anglin, M. D. (2000). Longitudinal patterns of drug use and treatment participation: Findings from the 5-year follow-up of DATOS. In College on Problems of Drug Dependence: Abstracts of the 62nd Annual Scientific Meeting, San Juan, Puerto Rico (p. 69). Philadelphia, PA: Temple University & College on Problems of Drug Dependence, Inc.

Hser, Y. I., Hoffman, V., Grella, C. E., & Anglin, M. D. (2001). A 33-year follow-up of narcotics addicts. Archives of General Psychiatry, 58, 503-508.

Hser, Y. I., Maglione, M., & Boyle, K. (1999). Validity of self-report of drug use among STD patients, ER patients, and arrestees. American Journal of Drug and Alcohol Abuse, 25(1), 81-91.

Hubbard, R. L., Craddock, G., & Anderson, J. (2001). Overview of 5-year follow-up outcomes in the Drug Abuse Treatment Outcome Studies (DATOS). Unpublished manuscript, Raleigh, NC.

McGlothlin, W. H., Anglin, M. D., & Wilson, B. D. (1977). An evaluation of the California civil addict program (DHEW Publication No. ADM 78-558, NIDA Services Research Monograph Series). Washington, DC: U.S. Government Printing Office.

Moos, R. H., Brennan, P. L., & Mertens, J. R. (1994). Mortality rates and predictors of mortality among late-middle-aged and older substance abuse patients. Alcoholism: Clinical and Experimental Research, 18(1), 187-195.

National Center for Health Statistics, Centers for Disease Control and Prevention. (2001). Classifications of diseases and functioning & disability: Home page. Retrieved December 21, 2001, from http://www.cdc.gov/nchs/icd9.htm

Nurco, D. N., Bonito, A. J., Lerner, M., & Balter, M. B. (1975). Studying addicts over time: Methodology and preliminary findings. American Journal of Drug and Alcohol Abuse, 2(2), 183-196.

Office of Applied Studies. (1998). Services Research Outcomes Study (SROS) (DHHS Publication No. SMA 98-3177, Analytic Series A-5). Rockville, MD: Substance Abuse and Mental Health Services Administration.

Vaillant, G. E. (1996). A long-term follow-up of male alcohol abuse. Archives of General Psychiatry, 53, 243-249.

________

* To whom correspondence should be sent at 1640 South Sepulveda Boulevard, Suite 200, Los Angeles, CA 90025. Telephone: 310-445-0874, x264. E-mail: yhser@ucla.edu.

1 Criteria for alcohol abuse and dependence are defined in the 3rd edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III) published by the American Psychiatric Association (APA, 1987).

2 Based on the International Classification of Diseases, Ninth Revision (ICD-9) (National Center for Health Statistics [NCHS], 2001).

Top Of PageTable Of Contents

This page was last updated on June 16, 2008.