Month: April 2023

RESEARCH ARTICLE Open Access

Trends in substance use admissions among older adults Sumedha Chhatre1*, Ratna Cook2, Eshita Mallik3 and Ravishankar Jayadevappa2,4,5,6,7

Abstract

Background: Substance abuse is a growing, but mostly silent, epidemic among older adults. We sought to analyze the trends in admissions for substance abuse treatment among older adults (aged 55 and older).

Methods: Treatment Episode Data Set – Admissions (TEDS-A) for period between 2000 and 2012 was used. The trends in admission for primary substances, demographic attributes, characteristics of substance abused and type of admission were analyzed.

Results: While total number of substance abuse treatment admissions between 2000 and 2012 changed slightly, proportion attributable to older adults increased from 3.4% to 7.0%. Substantial changes in the demographic, substance use pattern, and treatment characteristics for the older adult admissions were noted. Majority of the admissions were for alcohol as the primary substance. However there was a decreasing trend in this proportion (77% to 64%). The proportion of admissions for following primary substances showed increase: cocaine/crack, marijuana/hashish, heroin, non-prescription methadone, and other opiates and synthetics. Also, admissions for older adults increased between 2000 and 2012 for African Americans (21% to 28%), females (20% to 24%), high school graduates (63% to 75%), homeless (15% to 19%), unemployed (77% to 84%), and those with psychiatric problems (17% to 32%).The proportion of admissions with prior history of substance abuse treatment increased from 39% to 46% and there was an increase in the admissions where more than one problem substance was reported. Ambulatory setting continued to be the most frequent treatment setting, and individual (including self-referral) was the most common referral source. The use of medication assisted therapy remained low over the years (7% – 9%).

Conclusions: The changing demographic and substance use pattern of older adults implies that a wide array of psychological, social, and physiological needs will arise. Integrated, multidisciplinary and tailored policies for prevention and treatment are necessary to address the growing epidemic of substance abuse in older adults.

Keywords: Substance abuse, Older adults, Treatment episode data set – admissions (TEDS-A), Trends in admission

Background Substance abuse among older adults is one of the fastest growing health problems in the US [1–5]. The changing demographic composition of the older adult population in the US affects not only the prevalence of substance abuse, but also the need for a variety of services, includ- ing treatment. It is estimated that the number of older adults who will need treatment for substance abuse will increase from 1.7 million in 2000–2001 to 4.4 million in 2020 [3, 6]. This increase is partially attributed to the

aging baby boomer population who has had more expos- ure to drugs, alcohol and tobacco from a younger age, which is reported to be a risk factor for use and abuse of these substances in later years [6–8] . The use of illicit drugs among older adults appears to be increasing. A study showed that the use of illicit drugs among adults age 50–59 almost doubled between 2002 and 2007 (5.1% to 9.4%). Also, of the adults aged 50–59 who were using illicit drugs in 2007, almost 90% had started using them before age 30. This implies lifelong nature of illicit drug use [9]. Analysis of the 2008 data from the Drug Abuse Warning Network surveillance system showed that of the 1.1 million emergency department episodes for ad- verse drug reactions, 61% were for persons aged 65 or

* Correspondence: [email protected] 1Department of Psychiatry, Perelman School of Medicine, University of Pennsylvania, 3535 Market Street, Suite 4051, Philadelphia, Pennsylvania 19104, USA Full list of author information is available at the end of the article

© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Chhatre et al. BMC Health Services Research (2017) 17:584 DOI 10.1186/s12913-017-2538-z

older. Also, almost 25% these episodes were due to ad- verse reactions to central nervous system drugs [10]. Al- though limited, research indicates important racial and ethnic differences in the prevalence of substance abuse in older adults. Among persons aged 65–74, being white, male, and divorced or widowed was associated with higher odds of lifetime alcohol use disorder [11]. Being African American or Hispanic was one of the several factors associated with sub-threshold alcohol depend- ence in the past year [12]. In one study, African Ameri- cans aged 55 and older were reported to have higher prevalence as well as higher rates of treatment admis- sions for illicit drugs such as cocaine [13]. Important gender differences in the older persons with substance abuse are also noted. Women make-up a larger portion of the older population, especially among those aged 85 and older. The pattern of substance use in older women is different than that of older men. Research indicates that binge drinking in women aged 65 and older is lower compared to their male counterparts [14]. Also, older women had lower rates for alcohol dependence or abuse, drug dependence or abuse or both conditions, and lower past-year use of illicit drugs, compared to older men (1.4% vs. 2.2% in 2010) [10]. Even as the number of older adults with substance

abuse is on the rise, substance abuse is often undetected and undertreated in this population [15, 16]. Due to the stigma attached to substance abuse, elderly patients may not report this issue [17, 18]. Therefore, the true preva- lence of substance abuse in this population remains un- known. In addition, providers are often too busy or may confuse the symptoms of substance abuse disorders with other co-morbidities, age-related changes or reactions to stressful life situations [17, 19]. Number of co- morbidities increase with age and presence of substance abuse can lead to worsening of medical consequences and outcomes of care [19]. Of the total spending for substance abuse disorder treatments, a substantial share is borne by public sources: Medicare, Medicaid, local, state and federal governments. For example, in 2009, 69% of spending on substance abuse treatment came from public sources [20]. One study reported that com- pared to younger adults, the proportion of older adults seeking treatment for illicit drugs abuse for the first time is on the rise [2]. Thus, a rise in substance abuse among older persons

coupled with the aging of the US population has strong implications for treatment demands on the healthcare system. Though the definition of an ‘older adult’ may vary slightly, there is consensus that interaction of age related changes (physiological, psychological, functional or social) and substance abuse is detrimental to the well- being and exerts significant burden on the healthcare system. Following the criteria used by several prior

studies, in this study we define ‘older adults’ as those aged 55 or older [2, 21–23]. Information about trends in substance abuse pattern among older adults is essential for developing appropriate preventive and treatment pol- icies and for resource planning. Objective of this study was to analyze the national trends in admissions of older adults (aged ≥55 years) to publically funded substance abuse treatment facilities between 2000 and 2012. We analyzed the demographic attributes, characteristics of the substance abused, age at first use and type of admis- sion for the cohorts of older adult admissions between 2000 and 2012.

Methods Data source We used the Treatment Episode Data Set – Admissions (TEDS-A), an administrative public use data system that is maintained and sponsored by the Center for Behav- ioral Health Statistics and Quality at the Substance Abuse and Mental Health Services Administration (SAMHSA) [24]. All public and private substance abuse treatment facilities that receive public funds are required to report the information about annual flow of admis- sions via state funding agency to TEDS-A. In TEDS-A, the unit is analysis is an admission. This database also includes information on demographic characteristics (age, race, gender, employment, education, pregnancy, veteran status, health insurance), substance abuse behav- ior (type of substance, mode of use, frequency of usage, age at first use), treatment characteristics (referral source, prior treatment, service setting), geographic in- formation (region, division), and presence of psychiatric diagnosis at each admission.

Data analysis Our analysis included all admissions to the publically funded substance abuse treatment programs between 2000 and 2012 for persons who were aged 55 years or older at the time of admission. The unit of analysis is an admission and a person can have multiple admissions in a year. However, it is not possible to identify an individ- ual person and thus dependence of observations cannot be adjusted for. Given the large sample size, a stringent criteria of p < .0001 was adopted for determining statis- tical significance of Chisq tests. Substance use treatment admissions among older adults as a proportion of total admissions were compared over time. We also compared the trends in admission for following primary sub- stances: alcohol, cocaine/crack, heroin, marijuana/hash- ish, nonprescription methadone, other opiates and synthetics, methamphetamine, benzodiazepines, Phen- cyclidine, other hallucinogens, other amphetamines, other stimulants, other non-benzodiazepine, other non-

Chhatre et al. BMC Health Services Research (2017) 17:584 Page 2 of 8

barbiturates sedatives or hypnotics, inhalants, over-the- counter medications and other. In addition to demographic characteristics, sub-

stance characteristics, including the number of sub- stances abused at the time of admission, service treatment setting, referral source, number of prior treatment admission episodes, and use medication assisted therapy were analyzed for cohorts of older adult admissions. Finally, we analyzed the trend in type of substance abuse, and age at first initiation for those with no prior treatment admissions vs. those with at least one prior treatment admission.

Study results Of the total admissions to publically funded substance abuse treatment programs in year 2000, 3.4% (n = 60,112) were for older adults. There was a steady increase of this proportion over time, and in 2012, ad- missions for older adults accounted for 7.0% (n = 121,015) of all admissions. At the same time, the total number of admissions for substance abuse com- bined for all age-groups changed only slightly.

Demographic characteristics of cohorts of older adult admissions between 2000 and 2012 are presented in Table 1. Overall, most admissions from 2000 to 2012 were among non-Hispanic white, male, unmarried, high school graduates, unemployed and those with housing. However, some of these variables showed changes over time. For example, admissions for older adults increased between 2000 and 2012 for African Americans (21% to 28%), females (20% to 24%), high school graduates (63% to 75%), unmarried (79% to 84%), homeless (15% to 19%), unemployed (77% to 84%), and those with psychi- atric problems (17% to 32%). In Table 2, we present the type of substance that

caused the treatment admission and other characteristics of the admission. Among our cohorts of older adult ad- missions, majority of the admissions were for alcohol as the primary substance. However, there was a decreasing trend in this proportion from 2000 to 2012 (77% to 64%). On the other hand, proportion of admissions for following primary substances showed steep increase be- tween 2000 and 2012: cocaine/crack (63% increase), marijuana/hashish (150% increase), heroin (26%

Table 1 Demographics of substance use treatment admissions in older adults, TEDS-A 2000–2012

Characteristic 2000 (N = 60,112) n (%)

2004 (N = 69,310) n (%)

2008 (N = 104,431) n (%)

2012 (N = 121,015) n (%)

Gender*

Male 48,155(80.4) 54,320(78.4) 80,344(76.95) 91,255(75.54)

Female 11,742(19.6) 14,963(21.6) 24,071(23.05) 29,546(24.46)

Race/ethnicity*

White non-Hispanic 35,071(58.34) 39,219(56.58) 57,669(55.22) 67,068(55.42)

Black Non-Hispanic 12,874(21.42) 16,467(23.76) 27,630(26.46) 34,258(28.31)

Hispanic 9835(16.36) 10,973(15.83) 14,743(14.12) 14,371(11.88)

Other 2332(3.88) 2651(3.2) 4389(4.20) 5318(4.39)

Marital status*

Married 12,384(20.6) 14,017(20.22) 18,673(17.88) 19,748(16.32)

Not married 47,728(79.4) 55,293(79.78) 85,758(82.12) 101,267(83.68)

Education*a

Completed high school 36,545(62.99) 46,759(70.22) 75,480(73.98) 88,641(74.8)

Did not complete high school 21,470(37.01) 19,829(29.78) 26,546(26.02) 29,857(25.2)

Employment*b

Employed 12,959(22.97) 13,999(21.37) 20,756(20.15) 18,326(15.55)

Not employed 43,465(77.03) 51,499(78.63) 82,261(79.85) 99,544(84.45)

Living arrangement*c

Homeless 6844(14.55) 9689(16.93) 17,441(17.48) 22,344(18.69)

Not homeless 40,201(85.45) 47,536(83.07) 82,334(82.52) 97,238(81.31)

Other psychiatric illness*d

No 31,000(83.06) 31,486(75.59) 49,887(70.57) 57,743(67.62)

Yes 6321(16.94) 10,169(24.41) 20,805(29.43) 27,655(32.38)

*p-value <0.0001; a 2.7% missing; b 3% missing; c 9.75% missing; d 34% missing

Chhatre et al. BMC Health Services Research (2017) 17:584 Page 3 of 8

increase), non-prescription methadone (200% increase), other opiates and synthetics (221% increase), and benzo- diazepines (67% increase) (Table 2 and Fig. 1). For the base year (i.e., year 2000), the proportion of admissions for the following substance was under 1% and therefore not reported in Table 2: nonprescription methadone, methamphetamine, benzodiazepines, Phencyclidine, other hallucinogens, other amphetamines, other stimu- lants, other non-benzodiazepine, other non-barbiturates sedatives or hypnotics, inhalants, over-the-counter medi- cations and other. The proportion of admissions for poly-substances grew over time (20% in 2000 vs. 38% in 2012). Poly-substance is defined as use of secondary and/or tertiary problem substances, in addition to the primary substance as reported at the time of admission. Table 2 also presents the treatment characteristics of

the cohorts of older adult admissions between 2000 and 2012. The majority of the admissions were to

ambulatory setting (54%) and individual (including self- referral) was the most common referral source (40% – 43%). The use of medication assisted therapy remained low over the years (7% – 9%). The number of Admissions with prior history of substance abuse treatment in- creased from 39% to 46%. In year 2000, admissions where alcohol was the only substance reported accounted for two-third of all older adult admission. However, by 2012, less than half of all admissions were for alcohol only. At the same time, the proportion of ad- missions for other drugs only and alcohol plus other drugs increased between 2000 and 2012. In Table 3, we present the comparison of substance

abuse type across two sub-groups: those without prior admissions vs. those with at least one prior admission. Among those older adults without prior admissions, the proportion of admissions for alcohol only declined (70% in 2000 vs. 52% in 2012). On the other hand, the

Table 2 Substance use and treatment characteristics across substance use treatment admission episodes in older adults, TEDS-A 2000–2012 Characteristic 2000 N = 60,112

n (%) 2004 N = 69,310 n (%)

2008 N = 104,431 n (%)

2012 N = 121,015 n (%)

Number of substances*

1 46,227(76.9) 48,308(69.7) 67,689(64.82) 72,934(60.27)

2 8843(14.71) 13,563(19.57) 24,655(23.61) 32,791(27.1)

3 3349(5.57) 5516(7.96) 9933(9.51) 13,839(11.44)

Primary substance problem*

Alcohol 45,527(77.05) 47,185(69.8) 67,247(64.61) 78,003(64.68)

Heroin 6912(11.7) 9395(13.9) 15,802(15.18) 17,896(14.84)

Cocaine/crack 2865(4.85) 5112(7.56) 9394(9.03) 9449(7.84)

Other opiates and synthetics 799(1.35) 1696(2.51) 3580(3.44) 5470(4.54)

Marijuana/hashish 721(1.22) 1270(1.88) 2561(2.46) 3624(3.01)

Any substance use type

Alcohol only 39,675 (66.0) 9018 38,477 (55.51) 51,585 (49.40) 54,961 (45.42)

Other drug only (15.0) 14,369 (20.73) 25,198 (24.13) 29,680 (24.53)

Alcohol and other drug 9726 (16.18) 14,541 (20.98) 25,494 (24.41) 34,923 (28.86)

Service setting

Detox 19,463(32.38) 20,048 (30.17) 27,759 (26.59) 11,096 (30.92)

Rehab 7659(12.74) 9398 (13.57) 15,803 (15.14 18,162 (15.01)

Ambulatory 32,981 (54.87) 39,847 (57.51) 60,849 (58.28) 65,425 (54.05)

Referral source*

Individual (includes self-referral) 23,505(40.62) 27,433(40.73) 41,307(40.37) 51,308(43.27)

Healthcare provider (Alcohol/drug abuse /other) 14,366 (24.83) 15,225 (22.6) 22,642 (22.13) 25,021(21.1)

School (educational)/Employer/EAP 791(1.37) 892(1.33) 995(0.97) 848(0.71)

Other community referral 3761(6.5) 5862(8.7) 9887(9.66) 12,823(10.81)

Criminal justice 15,438(25.68) 17,943(25.89) 27,495(26.33) 28,584(23.62)

Number of prior episodes*a

0 20,503(42.98) 22,573(41.45) 36,092(39.85) 36,584(35.18)

≥ 1 27,198 (57.02) 31,888 (58.55) 54,489(60.15) 67,410(64.82)

Medication assisted therapy used*b Yes 4541(7.85) 5394(8.41) 9520(9.76) 10,792(9.35)

*p-value <0.0001; a 17.6% missing; b 5.6% missing

Chhatre et al. BMC Health Services Research (2017) 17:584 Page 4 of 8

proportion of admissions for other drugs only and those for alcohol plus other drugs almost doubled between 2000 and 2012. A similar pattern was observed for those with at least one prior admission. Age at first initiation also showed comparable trends between these two sub- groups (data not shown). When the admission was for alcohol only or for alcohol with drug, more than 80% re- ported age at first use as younger than 25%. However, when the admission was for other drugs only, only half reported first use at age younger than 25 years thus indi- cating continued initiation of drug use over the adult years, including 55 or older.

Discussion Substance abuse is an important psychosocial comorbid- ity in older adults and our results add to the growing body of evidence that the magnitude of substance abuse disorders among older adults is escalating [1–3, 5, 6, 22, 25, 26]. Longitudinal data from TEDS-A demonstrate that while the total number of admissions to the

publically funded substance abuse treatment programs have stayed almost constant between 2000 and 2012, there is an increase in the proportion of admissions at- tributable to older adults. Our results also show that ini- tiation of drug use is spread over the lifespan of the older adult as opposed to first use of alcohol which hap- pened mostly prior to age 25. Socio-cultural factors ap- pear to have a role in the observed variations in prevalence, type of substance abused, treatment charac- teristics and outcomes for substance abuse. Thus, the issue of substance abuse among older adults may be viewed in a broad sociocultural framework. Several demographic and service related factors may

have contributed to the changes in demographic com- position and treatment characteristics of the substance abusing older adults. One important demographic factor is the aging baby boomer cohort (those born between 1946 and 1964). The baby boomer cohort turns 55 years old between 2001 and 2019, and 65 years old between 2011 and 2029. In addition to being larger in numbers,

Table 3 Substance abuse type for older adults by number of prior admissions

Substance abuse type** No prior admission * One or more prior admissions*

2000 (n = 20,503)

2004 (n = 22,573)

2008 (n = 36,092)

2012 (n = 36,584)

2000 (n = 39,609)

2004 (n = 46,737)

2008 (n = 68,339)

2012 (n = 84,431)

Alcohol only 14,376 13,497 19,868 19,155 25,299 24,530 31,717 35,906

(70.1) (61.8) (n = 55.1) (n = 52.5) (63.9) (52.5) (46.4) (42.4)

Other drugs only 2478 4005 7919 8719 6540 10,364 17,279 20,961

(12.1) (17.7) (21.9) (23.8) (16.5) (22.2) (25.3) (24.8)

Alcohol and other drugs 2615 3702 7136 8260 7111 10,839 18,358 26,663

(12.7) (16.4) (19.8) (22.6) (17.9) (23.2) (26.9) (31.6)

None 1034 919 1169 450 659 1004 985 1001

(5.0) (4.1) (3.2) (1.2) (1.7) (2.1) (1.4) (1.2)

**Primary, secondary or tertiary substance of abuse reported at the time of admission *p-value <0.0001

Fig. 1 Percent of admissions with specific substance as the primary substance of abuse of those aged ≥ 55

Chhatre et al. BMC Health Services Research (2017) 17:584 Page 5 of 8

this cohort also has higher prevalence of lifetime sub- stance use, compared to earlier older cohorts [6, 27]. In addition, a history of alcohol abuse increases the risk of substance use in late life [26]. Thus, some of the ob- served growth in number of older adult admissions to substance abuse treatment program may be a reflection of aging of the baby boomer cohort. Although alcohol remains the top primary reason for admission among older adults, the number of admissions where alcohol was either the primary or the only substance abused have decreased between 2000 and 2012. On the other hand, there was a significant increase in the proportion of admissions for drug use only or for combined drug and alcohol use. Source of referral also offers some interesting insight

into the changing composition of older adult admissions to substance abuse treatment programs. Referrals from other community sources increased between 2000 and 2012. However, overall referrals by healthcare providers declined over the study period, 24% in 2000 vs. 21% in 2012. Also, the overall referrals made by criminal justice system declined over time. This result is similar to the one reported by a study of TEDS-A for the period be- tween 1992 to 2005 [22]. While increase in community referrals and individual referrals suggests better aware- ness and access to substance abuse treatments, the de- crease in referrals from providers is indicative of lost- opportunity for screening and referrals [21]. Time pres- sure, lack of training and mistaking substance abuse symptoms for those associated with normal aging may be some of the reasons for the decrease in referrals from providers. Medication assisted therapy is commonly used in

treatment for alcohol and opioid-related addictions. However, despite research demonstrating the effective- ness of medication assisted therapy as an evidence-based practice for substance abuse, such treatment remains underutilized. We observed only 7.9–9.8% of total ad- missions in older adults reporting medication assisted opioid therapy as part of the treatment plan. A separate analysis (data not reported) showed that for admissions where other drugs (and not alcohol) was the substances abused, 43% of admission had medication assisted opioid therapy as part of the treatment plan, however by year 2012, this proportion decreased to 31%. Additionally, medication assisted therapy was reported by a very small proportion of admissions that were for alcohol plus other drug abuse. Another study has reported that less than one-half of the 2.5 million Americans aged 12 or older who abused or were dependent on opioids in 2013 received medication assisted therapy [28]. In our study, admissions for older adults where primary substance was opiates and other synthetics increased by 221% be- tween 2000 and 2012. It is not surprising that opiates,

which are the most commonly prescribed drug class in the US, have shown the most increase in admissions re- lated to substance abuse. The nearly 9-fold increase in opioid prescriptions from office based medical visits by older adults that occurred between 1995 and 2010 sug- gests that physicians have pursued greater pain control in this group [20]. Additionally it has been shown that overdose deaths involving opioid analgesics now exceed deaths involving heroin and cocaine combined [5]. We note certain limitations to our study that are intrinsic

to the TEDS-A data. First and foremost, the unit of analysis in TEDS-A data is an admission, and not a person. It is dif- ficult to determine if the increase in number of admissions is due to an increase in the number of unique older adults seeking treatment, or is a reflection of multiple admissions made by a smaller group of older adults. To some extent, we addressed this limitation by analyzing the number of prior visits in order to isolate those new to the system (no prior admissions) versus those who have repeat admissions (at least one prior admission). Secondly, TEDS does not in- clude non-publically funded substance abuse treatment programs or data from the Department of Veterans Ad- ministration. The demographic and substance abuse profile of those not seeking care in a publically funded substance abuse program may be different than that observed in TEDS-A, and this may affect the generalizability of the re- sults. As the admissions reported in TEDS-A are from pub- lically funded substance use programs, they are affected by peripheral factors such as the availability of funds, target groups, and state policies. Additionally, the self-reported in- formation in TEDS has potential for memory and/or per- sonal biases. Finally, completeness of data reporting may differ by state and may lead to variation in the magnitude of available data. Thus, interpretations of our results must be made within the context of these data limitations.

Policy implications Our findings have several policy implications. First, our results provide an indication of service needs that is es- sential for planning purposes. The healthcare system must be prepared to treat a large number of older adults with substance abuse resulting from a growth in the number of older adults coupled with the growing sub- stance abuse epidemic in this population. The changing composition of this cohort implies a wide array of psy- chological, social, and physiological needs that must be addressed in the coming years. For example, as older adults often come in contact with their primary care or other healthcare providers, it is important that these providers are able to distinguish substance abuse prob- lems from physical or mental health problems, and refer the patients to treatment as needed. Given the chal- lenges in diagnosing substance abuse in aging elderly, policies that support geriatric education and training for

Chhatre et al. BMC Health Services Research (2017) 17:584 Page 6 of 8

healthcare providers and others working with the aging population (for e.g. aging service providers) may be beneficial. Additionally, the type of substance abused and treatment setting presents specific challenges to publically funded treatment programs, including Medi- care and Medicaid. The Affordable Care Act includes many provisions to improve and expand treatment for people with substance abuse disorders, and also expands the Medicaid programs in certain states. Therefore, the issue of substance abuse among older adults can be ef- fectively addressed by integrated and multidisciplinary collaboration among the treatment community and other service systems for aging adults.

Conclusions Substance abuse among older adults is a serious issue and its magnitude will grow with the aging of the baby boomer cohort. Even though the total number of sub- stance abuse admissions between 2000 and 2012 remained mostly unchanged, the proportion of admis- sions attributable to older adults increased more than two-fold. While alcohol still remains the most frequent reason for the admission to substance abuse treatment, this proportion is declining. On the other hand, cocaine and heroin related admissions (i.e., where cocaine or heroin was the client’s primary substance problem) are on the rise in the older adult population. Substance abuse translates into treatment need, may

affect health outcomes and complicate the treatment of other comorbid conditions among older persons. Major- ity of the substance abu