Drug and Alcohol Dependence 142 (2014) 161–167
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Full length article
The association between nonmedical use of prescription medication status and change in health-related quality of life: Results from a Nationally Representative Survey Ty S. Schepis a,∗ , Jahn K. Hakes b a b
Department of Psychology, Texas State University, 601 University Drive, San Marcos, TX 78666, United States Center for Administrative Records Research and Applications, U.S. Census Bureau, Suitland, MD, United States
a r t i c l e
i n f o
Article history: Received 5 March 2014 Received in revised form 4 June 2014 Accepted 7 June 2014 Available online 19 June 2014 Keywords: Nonmedical prescription use Prescription misuse Opioid Tranquilizer Stimulant Sedative SF-12 Health-related quality of life
a b s t r a c t Background: Nonmedical use of prescription medication (NUPM) is associated cross-sectionally with a host of medical and psychosocial consequences. Few studies, however, have examined longitudinal outcomes based on NUPM indicators. This study aimed to address this gap by examining change in health-related quality of life as a function of NUPM status. Methods: Data are from waves 1 and 2 of the National Epidemiological Survey on Alcoholism and Related Conditions (NESARC) a household-based, nationally-representative survey of the US population. 34,653 participants who completed both NESARC waves were included in analyses. The primary outcome measure was the 12-item Short Form Health Survey (SF-12), with history of NUPM of opioids, tranquilizer/sedatives and stimulants (examined separately) at wave 1 and any NUPM between waves 1 and 2 used to group participants. Sociodemographic characteristics were used as control variables. Results: Across medication classes, results indicated that individuals who initiated NUPM between waves (initiators) had greater declines or smaller increases on many SF-12 scales, when compared to other groups. Individuals with a history of NUPM at wave 1 but no use between waves (quitters) and never users generally had the best outcomes in terms of change in SF-12 scales, with quitters making larger gains (or smaller losses) in mental health-related quality of life. Persistent users were generally intermediate between initiators and quitters or never users. Conclusions: These data reinforce the importance of preventing NUPM initiation and of promoting NUPM cessation, highlighting the need for greater use of NUPM-related public health interventions. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The nonmedical use of prescription medications (NUPM) in the United States has been described as a public health concern (Garnier et al., 2009; Schepis and Hakes, 2013) or even an epidemic (Centers for Disease Control and Prevention, [CDC], 2012), with rates of NUPM and NUPM-related substance use disorder (SUD) having increased notably over the past 20 years (Martins et al., 2010; McCabe et al., 2007). While rates have leveled off somewhat, rates of opioid NUPM trail only those of alcohol, tobacco and marijuana (Substance Abuse and Mental Health Services Administration [SAMHSA], 2012); when combined with NUPM from other classes of medication (e.g., tranquilizers, sedatives and stimulants), the
∗ Corresponding author. Tel.: +1 512 245 6805; fax: +1 512 245 3153. E-mail address: [email protected] (T.S. Schepis). http://dx.doi.org/10.1016/j.drugalcdep.2014.06.009 0376-8716/© 2014 Elsevier Ireland Ltd. All rights reserved.
total past year prevalence of NUPM is 5.6%, which is larger than the pooled past year prevalence rates (3.9%) of cocaine, methamphetamine, heroin or any hallucinogen use (SAMHSA, 2012). With increases in NUPM rates, rates of NUPM-related emergency department utilization, substance use treatment and overdose have increased dramatically (CDC, 2012; Manchikanti and Singh, 2008; SAMHSA, 2009, 2010). Moreover, substantial crosssectional research has associated NUPM with serious psychosocial and neurobehavioral consequences, including psychiatric illness (Goodwin and Hasin, 2002), suicidality (Kim et al., 2012; Kuramoto et al., 2012), other substance use (Boyd et al., 2006b; Schepis and Krishnan-Sarin, 2008), poorer academic performance (McCabe et al., 2005), unemployment (Simoni-Wastila and Strickler, 2004) and poorer self-reported health (Havens et al., 2011). Longitudinal research indicates that NUPM is associated with increased incidence and recurrence of psychiatric disorders (Martins et al., 2012; Schepis and Hakes, 2011) and that psychiatric disorders may
162
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
increase the likelihood of NUPM and NUPM-related SUD (Martins et al., 2012). Longitudinal data also indicate that certain physical conditions increase the risk of NUPM-related SUD at a 3-year follow-up (Katz et al., 2013). Despite these initial efforts, more longitudinal work is needed to examine the processes that promote NUPM and the consequences of NUPM. To illustrate, no research has examined how NUPM course longitudinally affects health-related quality of life. Health-related quality of life is the individual’s perception of his or her physical, mental and social functioning and how such functioning affects his or her well-being and daily activities (Gonzalez-Saiz et al., 2009). The evidence across substance users, with the greatest number of investigations in alcohol and opioid users, indicates that substance users have poorer health-related quality of life than non-users, with the worst health-related quality of life in those with more frequent or severe use (Gonzalez-Saiz et al., 2009; Ugochukwu et al., 2013). In turn, abstinence from substance use is associated with improvements in health-related quality of life (Best et al., 2013; Ugochukwu et al., 2013). As noted above, NUPM increases the risk of a psychiatric diagnosis over a 3-year follow-up (Martins et al., 2012; Schepis and Hakes, 2011), but no published work has investigated the longitudinal impact of NUPM on a more continuous measure of mental health-related quality of life. While use of categorical diagnoses provides important data about NUPM-related mental health outcomes, diagnostic categorization may miss larger trends toward declining mental health-related quality of life in nonmedical users, especially if such declines are not severe enough in many to warrant a diagnosis. Examination of the longitudinal change in self-reported physical and mental health-related quality of life could provide important data on the consequences of NUPM. 1.1. Aims and hypotheses This work aims to accomplish this goal through the use of two waves of data from the National Epidemiological Survey on Alcohol and Related Conditions (NESARC). Both waves of the NESARC contain comprehensive assessments of NUPM using opioids, tranquilizers, stimulants and sedatives and the 12-item Short Form Health Survey (SF-12), a validated measure of physical and mental health-related quality of life (Ware et al., 1996). The primary aim was to evaluate changes in SF-12 scores over the 3-year period between NESARC waves by NUPM course: never users (denied any NUPM at both waves), initiators (denied lifetime NUPM at wave 1, but endorsed NUPM between waves), quitters (endorsed past year NUPM at wave 1, but denied NUPM between waves) and persistent users (endorsed lifetime NUPM at wave 1 and between waves). Analyses were performed separately by medication group, with tranquilizers and sedatives combined in a single group; comparisons examined differences in SF-12 scale and subscale change scores between participant subgroups. Across medication class, we hypothesized that never users would have the largest gains or smallest declines in SF-12 scores over the follow-up, followed by quitters and initiators. We also hypothesized that persistent users would have the worst outcomes, as signified by the largest decreases or smallest gains on the SF-12. 2. Methods The NESARC is a longitudinal survey funded by the National Institute on Alcohol Abuse and Alcoholism (NIAAA), with two available data waves (wave 1: 2001–2002; wave 2: 2004–2005). The NESARC aims to provide a representative sample of the US adult population for analyses of substance use behaviors and correlates. The 2000–2001 Census Supplementary Survey structured the sampling, which is described elsewhere (Grant et al., 2003b; Grant and Kaplan, 2005). The NESARC includes weighting to produce nationally representative data, and to adjust for survey design, young adult oversampling, and non-response at both wave 1 and wave 2 (among wave 1 participants). The US Census Bureau and the US Office
of Management and Budget approved the NESARC protocol, and the Texas State University IRB approved this investigation. At wave 1, 43,093 individuals participated with 39,959 eligible to participate in wave 2. Of those eligible, 34,653 consented to participate in wave 2. The wave 1 response rate was 81.2%, and the wave 2 response rate was 86.7%, with a total response rate of 70.2% (Grant et al., 2003b; Grant and Kaplan, 2005). The weighted NESARC sample is 52% female, 71% Caucasian, 12% Hispanic/Latino and 11% AfricanAmerican; 13% was under 25 years of age.
2.1. Measures 2.1.1. Lifetime nonmedical use of prescription medication (NUPM). In the NESARC, NUPM is defined as prescription use “without a prescription, in greater amounts, more often, or longer than prescribed, or for a reason other than a doctor said you should use them.” Lifetime NUPM and NUPM between waves 1 and 2 were assessed for four classes of medication: opioids, tranquilizers, stimulants and sedatives. Due to low base rates for sedative NUPM, similar pharmacological properties and motives for sedative and tranquilizer NUPM (Boyd et al., 2006a; Hertz and Knight, 2006), sedative and tranquilizer NUPM were pooled.
2.1.2. Frequency of NUPM. For participants who endorsed lifetime NUPM from a given medication class, follow-up questions are asked at wave 1 to determine lifetime maximum use frequency and at wave 2 to assess frequency of NUPM within the past year.
2.1.3. Substance Use Disorder (SUD) from NUPM. SUD diagnosis was obtained through the NIAAA Alcohol Use Disorder and Associated Disabilities Interview Schedule–DSM-IV Edition (AUDADIS-IV; Grant et al., 2003a, 1995). The AUDADIS-IV is a fully structured diagnostic interview that assesses symptoms of many DSMIV (American Psychiatric Association, 2000) disorders. The AUDADIS-IV appears to have good reliability and validity in assessing SUD (Grant et al., 2003a, 1995). At wave 1, lifetime SUD from NUPM was captured, and at wave 2, SUD from NUPM during the follow-up period was assessed.
2.1.4. Axis I diagnosis. Here, Axis I psychiatric outcomes assessed by the AUDADISIV include two depressive disorders (MDD or dysthymia), two bipolar disorders (bipolar I or II), six anxiety disorders (panic disorder with and without agoraphobia, agoraphobia without panic, social phobia, specific phobia or generalized anxiety disorder) and pathological gambling. Disorders were assessed at wave 1, were for past year diagnosis and were included in models as a dichotomous control variable.
2.1.5. Axis II diagnosis. Seven lifetime Axis II personality disorder (PD) diagnoses assessed at wave 1 and three PDs assessed at wave 2 were included in models as a dichotomous control variable. PDs assessed at wave 1 were antisocial, avoidant, dependent, obsessive-compulsive, paranoid, schizoid, and histrionic PD, with borderline, schizotypal and narcissistic PDs assessed at wave 2. The PDs assessed at wave 1 were only assessed at that timepoint, and the PDs assessed at wave 2 were only assessed at that point; in both cases only lifetime PD diagnosis was assessed.
2.1.6. 12-item Short-Form Health Survey, version 2 (SF-12). The SF-12 is a shorter version of the 36-item Short-Form Health Survey, with comparable reliability and validity indicators (Gandek et al., 1998; Ware et al., 1996). The SF-12 contains questions about physical and mental health-related quality of life, answered on a 5-point Likert-type scale. The SF-12 produces two summary scales, the physical component and mental component scales (PCS and MCS, respectively) and eight subscales. The physical functioning, role-physical, bodily pain and general health subscales compose the larger PCS scale, and the vitality, social functioning, role-emotional and mental health subscales compose the larger MCS scale. Sociodemographic variables assessed were age, gender, race/ethnicity, marital status, education level, household income, employment/full-time student status, and region of participant residence.
2.2. Subsamples Analyses split participants into 4 groups for each class of medication, based on NUPM status at wave 1 and status between waves. Participants who denied NUPM at both waves were grouped as never users, while those who denied lifetime NUPM at wave 1 but endorsed NUPM between waves were described as initiatiors. Participants who endorsed both lifetime NUPM at wave 1 and NUPM between waves were grouped as persistent users. The only exception to the lifetime NUPM classification was in quitters, who were classified as those endorsing NUPM in the year prior to wave 1 but no NUPM between waves. Quitters were so classified to isolate the effects of NUPM cessation, as using “lifetime quitters” who engaged in lifetime NUPM with no NUPM between waves would have included many individuals who ceased NUPM many years prior. Participants with lifetime NUPM but no past year NUPM and no NUPM between NESARC waves were excluded from analyses.
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
163
Table 1 Means of SF-12 score change from wave 1 to wave 2 by opioid nonmedical use group.
Physical Component Physical Functioning Role Physical Bodily Pain General Health Mental Component Vitality Social Functioning Role Emotional Mental Health
Never Users (1) (n = 32419; 95.8%)
Initiators (2) (n = 735; 2.2%)
Quitters (3) (n = 412; 2.2%)
Persistent Users (4) (n = 262; 0.8%)
Comparisons between groups, by group number
−0.64 (−0.70, −0.59) −0.90 (−0.96, −0.84) −1.47 (−1.54, −1.40) 0.43 (0.36, 0.50) −1.10 (−1.18, −1.01) −1.18 (−1.27, −1.09) −1.78 (−1.89, −1.67) −0.51 (−0.60, −0.42) −1.89 (−1.99, −1.79) −0.35 (−0.43, −0.26)
−1.43 (−1.72, −1.15) −1.41 (−1.67, −1.14) −2.45 (−2.84, −2.07) −1.10 (−1.74, −0.45) −1.51 (−1.91, −1.11) −2.19 (−2.74, −1.63) −2.41 (−2.88, −1.95) −2.10 (−2.82, −1.38) −2.91 (−3.40, −2.43) −1.14 (−1.70, −0.58)
0.84 (0.59, 1.09) −0.48 (−0.89, −0.08) 0.59 (0.13, 1.05) 2.06 (1.52, 2.59) 0.93 (0.64, 1.22) −0.11 (−0.53, 0.31) −0.60 (−0.93, −0.27) 0.55 (−0.05, 1.16) −0.98 (−1.69, −0.27) 0.97 (0.45, 1.48)
−1.38 (−1.82, −0.94) −1.77 (−2.09, −1.45) −3.28 (−3.81, −2.74) −0.45 (−1.20, 0.30) −0.82 (−1.68, 0.03) −2.50 (−3.41, −1.60) −3.36 (−3.98, −2.74) -1.63 (−2.49, −0.77) −3.12 (−4.19, −2.04) −1.63 (−2.63, −0.63)
3 > all; 1 > 2 1 > 2, 4 3 > all; 1 > 2, 4 3 > 2, 4; 1 > 2 3 > all 1, 3 > 2, 4 1, 3 > 4 1, 3 > 2, 4 1 > 2, 4 3 > 2, 4; 1 > 4
Notes: Means of change from wave 1 to wave 2 are listed with 95% confidence intervals in parentheses; n.s.d = no significant differences; the “Comparisons” column uses multivariate regression to compare groups, after controlling for age, gender, race/ethnicity, marital status, education, household income, employment/student status, region of participant residence and past year DSM-IV Axis I and lifetime Axis II psychopathology, with ˛ = 0.05.
2.3. Data analysis SF-12 mean scores for each medication class and each subgroup of participants were calculated, along with frequencies/percentages of sociodemographic characteristics for each subgroup. Primary analyses used design-based linear regression, comparing change in SF-12 score for each component scale and subscale between members of each subgroup. Analyses were run seperately by class of medication, and analyses controlled for the 8 sociodemographic variables listed above, past year (from wave 1) Axis I psychiatric diagnosis and lifetime Axis II diagnosis. Followup analyses examined three nonmedical use characteristics between initiators and persistent users: (1) past year nonmedical use, (2) past year weekly nonmedical use, and (3) SUD from nonmedical use in the period between wave 1 and wave 2. Followup analyses also examined potential differences in weekly nonmedical use at the lifetime maximum use level between quitters and persistent users. Finally, analyses compared initiators, quitters and persisitent users on rates of lifetime NUPM-related SUD. Follow-up analyses used Rao-Scott chi-square (Rao and Scott, 1984) to compare prevalence rates seperately by class of medication. All data were weighted, clustered on primary sampling units, and stratified appropriately. We included models only if they evidenced adequate fit. All analyses were performed in SAS, version 9.2 (Cary, NC).
3. Results 3.1. Subsample prevalence and sociodemographics The largest participant group was never users (opioids: 95.8%; tranquilizer/sedatives: 96.0%; stimulants: 99.1%), followed by initiators (opioids: 2.2%; tranquilizer/sedatives: 1.9%; stimulants: 0.4%) and quitters (opioids: 1.2%; tranquilizer/sedatives: 1.3%; stimulants: 0.4%). Persistent users were the smallest group (opioids: 0.8%; tranquilizer/sedatives: 0.8%; stimulants: 0.2%). Sociodemographic and NUPM characteristics for each group are included as Supplementary Material.1 Across medication classes, a higher proportion of females and non-Caucasian participants were never users, with the second highest proportion in the initiators group. Persistent users had the highest proportion of male and Caucasian participants. Otherwise, patterns were less consistent. Never users tended to be older and more likely to be married, with initiators and persistent users having higher proportions in the 18–25 cohort and the never married group. Persistent users and quitters were more likely to have household incomes below $40,000, though this did not clearly hold for tranquilizer/sedative users.
Table 1. For change in PCS score, quitters’ scores improved while all other subgroups’ scores declined, and never users had a smaller decline over the follow-up than opioid NUPM initiators. All PCS subscales evidenced significant subgroup differences: one, never users had smaller declines in physical functioning than initiators and persistent users; two, quitters evidenced improvements on the role physical subscale while all other subgroups declined, and never users had smaller declines than initiators or persistent users; three, quitters had better change outcomes on the bodily pain subscale than initiators or persistent users, and never users had superior follow-up outcomes than initiators; finally, quitters had improvements in general health between waves while all other subgroups evidenced declines. For change in MCS score and the social functioning subscale, opioid NUPM quitters and never users had significantly more adaptive changes in functioning than initiators or persistent users. Both quitters and never users had better longitudinal outcomes on the vitality and mental health subscales than persistent users; quitters were also superior to initiators in change in the mental health and social functioning subscales. Finally, never users had smaller declines between waves on the role emotional subscale than initiators or persistent users. 3.3. Change in SF-12 scores by tranquilizer/sedative nonmedical use group All means and 95% confidence intervals for SF-12 change scores by tranquilizer/sedative subgroup are listed in Table 2. The PCS and its subscales evidenced an inconsistent pattern of longitudinal results, with tranquilizer/sedative persistent users having significantly better change in PCS scores than quitters. Never users had smaller declines between waves in the physical functioning subscale than either initiators or quitters, and never users had smaller declines on the role-emotional subscale than initiators. No significant differences were found on the bodily pain or general health subscales. On the MCS, and the social functioning and the mental health subscales, never users and quitters had superior longitudinal outcomes than both initiators and persistent users. For the vitality and role-emotional subscales, the never user and quitter subgroups evidenced smaller declines in change scores than did initiators.
3.2. Change in SF-12 scores by opioid nonmedical use group 3.4. Change in SF-12 scores by stimulant nonmedical use group All means and 95% confidence intervals for SF-12 component and subscale change scores by opioid use subgroup are listed in
1 Supplementary material can be found by accessing the online version of this paper at http://dx.doi.org and by entering doi:. . ..
All means and 95% confidence intervals for SF-12 change scores by stimulant subgroup are listed in Table 3. Comparison of the stimulant subgroups demonstrated many fewer significant differences in change scores than those seen in the opioid or tranquilizer/sedative comparisons. Quitters had superior longitudinal
164
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
Table 2 Means of SF-12 score change from wave 1 to wave 2 by tranquilizer/sedative nonmedical use group.
Physical Component Physical Functioning Role Physical Bodily Pain General Health Mental Component Vitality Social Functioning Role Emotional Mental Health
Never Users (1) (n = 32232; 96.0%)
Initiators (2) (n = 653; 1.9%)
Quitters (3) (n = 427; 1.3%)
Persistent Users (4) (n = 252; 0.8%)
Comparisons between groups, by group number
−0.65 (−0.71, −0.59) −0.89 (−0.95, −0.83) −1.47 (−1.54, −1.41) 0.40 (0.33, 0.48) −1.14 (−1.22, −1.05) −1.19 (−1.28, −1.11) −1.77 (−1.87, −1.66) −0.54 (−0.62, −0.45) −1.90 (−2.00, −1.80) −0.36 (−0.44, −0.27)
−0.79 (−1.06, −0.53) −1.63 (−1.91, −1.36) −2.44 (−2.81, −2.07) −0.42 (−0.85, 0.003) −1.67 (−1.96, −1.38) −3.92 (−4.48, −3.35) −3.58 (−4.03, −3.13) −2.67 (−3.22, −2.12) −3.70 (−4.22, −3.17) −3.09 (−3.61, −2.56)
−1.22 (−1.55, −0.89) −1.58 (−1.96, −1.21) −1.93 (−2.19, −1.66) 0.24 (−0.19, 0.66) −0.24 (−0.60, 0.11) 0.29 (−0.07, 0.64) −0.53 (−0.97, −0.09) 0.48 (0.14, 0.82) −1.63 (−1.98, −1.28) 0.83 (0.47, 1.18)
0.61 (0.25, 0.98) −0.74 (−1.04, −0.44) −1.26 (−1.79, −0.72) 2.10 (0.90, 3.29) 0.04 (−0.61, 0.70) −2.60 (−3.89, −1.32) −2.22 (−2.92, −1.52) −2.03 (−3.03, −1.03) −2.15 (−3.57, −0.74) −1.49 (−2.64, −0.34)
4>3 1 > 2, 3 1>2 n.s.d. n.s.d. 1, 3 > 2, 4 1, 3 > 2 1, 3 > 2, 4 1, 3 > 2 1, 3 > 2, 4
Notes: Means of change from wave 1 to wave 2 are listed with 95% confidence intervals in parentheses; n.s.d = no significant differences; the “Comparisons” column uses multivariate regression to compare groups, after controlling for age, gender, race/ethnicity, marital status, education, household income, employment/student status, region of participant residence and past year DSM-IV Axis I and lifetime Axis II psychopathology, with ˛ = 0.05.
outcomes than never users on the PCS, and the change scores of quitters were superior to all groups on the general health change score. Other physical health-related quality of life subscales (e.g., physical functioning, role physical, or bodily pain) did not evidence significant subgroup differences in change scores. Similarly, the MCS scale, role emotional and mental health subscales did not indicate significant changes in scores between stimulant subgroups. Only two mental health-related quality of life subscales displayed significant differences: both the vitality and social functioning subscales indicated that never users and quitters had superior longitudinal outcomes over initiators. 3.5. Comparison of NUPM characteristics between use groups In order to further characterize the three groups engaged in lifetime NUPM (i.e., initiators, quitters and persistent users), their nonmedical use characteristics were compared. Across medication class, persistent users had higher rates of past year (to wave 2) NUPM and NUPM-related SUD diagnosis during the follow-up period than initiators. Conversely, past year rates of weekly NUPM did not differ between initiators and persistent users for any medication class. Rates of weekly NUPM at the respondent’s maximum level of NUPM were significantly higher for quitters than persistent users across medication classes. Finally, across medication classes, rates of lifetime NUPM-related SUD were highest in persistent users, followed by quitters. Initiators had the lowest rates of lifetime NUPM-related SUD, and all three groups were significantly different with one exception: comparing initiators and quitters for tranquilizer/sedative-related SUD. Prevalence rates with 95% confidence intervals for these NUPM characteristics and the Wald F-test output are in Table 4.
4. Discussion This investigation produced both expected (e.g., the relatively adaptive longitudinal outcomes of never users) and unexpected results (e.g., the superior longitudinal health-related quality of life outcomes of quitters and the particularly poor follow-up outcomes seen in initiators) in light of our stated hypotheses. Across medication classes, never users generally had superior outcomes to initiators and persistent users, as expected. Contrary to our expectations, though, quitters often had roughly equivalent outcomes to never users across the SF-12 scale change scores, with superior outcomes to initiators and persistent users, particularly on mental health-related quality-of-life scales. These patterns were much less apparent for the stimulants, but they largely held for the opioid and tranquilizer/sedative subgroups. Along with the finding that NUPM cessation was beneficial, the other key finding of this work was that NUPM initiation was associated with notably greater declines in health-related quality of life, relative to never users and quitters.
4.1. Differences based on NUPM subgroup While we expected that never users would have better changes in SF-12 scores over the three-year follow-up period than any other group, we found that quitters were roughly equivalent on SF-12 change scores. For the PCS scale and the physical component subscales, the change in health-related quality of life for quitters was superior to even never users for opioid NUPM. The equivalent longitudinal mental component profile of quitters to never users may suggest that most of these users would fall into what is called an “experimenter” group, using infrequently before ceasing use. Starting with the work of Shedler and Block (1990), many, but not all
Table 3 Change from wave 1 to wave 2 in SF-12 scores by stimulant nonmedical use group.
Physical Component Physical Functioning Role Physical Bodily Pain General Health Mental Component Vitality Social Functioning Role Emotional Mental Health
Never Users (1) (n = 33031; 99.1%)
Initiators (2) (n = 125; 0.4%)
Quitters (3) (n = 123; 0.4%)
Persistent Users (4) (n = 51; 0.2%)
Comparisons between groups, by group number
−0.66 (−0.71, −0.60) −0.94 (−1.00, −0.88) −1.49 (−1.56, −1.43) 0.42 (0.35, 0.50) −1.12 (−1.21, −1.04) −1.21 (−1.30, −1.12) −1.77 (−1.88, −1.67) −0.54 (−0.62, −0.46) −1.94 (−2.03, −1.84) −0.38 (−0.47, −0.29)
−1.40 (−2.68, −0.12) −1.32 (−2.54, −0.10) −1.98 (−2.93, −1.03) −0.92 (−3.42, 1.58) −2.11 (−3.48, −0.73) −3.38 (−4.30, −2.45) −5.21 (−5.80, −4.61) −3.95 (−4.87, −3.03) −2.39 (−2.86, −1.93) −1.49 (−2.25, −0.74)
1.76 (0.23, 3.29) 0.42 (−0.03, 0.87) 0.30 (−0.77, 1.37) 2.49 (0.57, 4.41) 2.07 (1.76, 2.37) −0.86 (−2.58, 0.85) −0.79 (−1.86, 0.28) 0.67 (0.15, 1.19) −1.42 (−2.72, −0.13) 0.35 (−1.13, 1.83)
−0.52 (−1.79, 0.74) −2.78 (−4.37, −1.20) −1.65 (−2.51, −0.80) 2.94 (1.09, 4.79) −2.24 (−2.35, −2.13) −2.54 (−6.80, 1.71) −2.09 (−2.47, −1.70) −2.36 (−10.04, 5.32) −1.57 (−1.76, −1.37) −2.75 (−6.27, 0.77)
3>2 n.s.d. n.s.d. n.s.d. 3 > all n.s.d. 1, 3 > 2 1, 3 > 2 n.s.d. n.s.d.
Notes: Means of change from wave 1 to wave 2 are listed with 95% confidence intervals in parentheses; n.s.d = no significant differences; the “Comparisons” column uses multivariate regression to compare groups, after controlling for age, gender, race/ethnicity, marital status, education, household income, employment/student status, region of participant residence and past year DSM-IV Axis I and lifetime Axis II psychopathology, with ˛ = 0.05.
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
165
Table 4 Subgroup nonmedical use characteristics examined separately by medication class. Opioids
Tranquilizer/Sedative
Stimulant
PY NM Use Initiators Persistent Users Comparison
67.1% (65.1–69.2%) 75.2% (71.8–78.7%) F(1, 35) = 15.09, p = .0004
65.5% (63.9–67.2%) 75.1% (71.6–78.6%) F(1, 28) = 25.34, p < .0001
65.2% (62.8–67.6%) 79.9% (76.2–83.5%) F(1, 5) = 81.09, p = .0003
PY Weekly NM Use Initiators Persistent Users Comparison
25.8% (24.0–27.6%) 20.2% (14.6–25.8%) F(1, 22) = 3.88, p = .062
23.2% (21.3–25.2%) 20.8% (16.9–24.7%) F(1, 18) = 1.07, p = .31
23.6% (22.0–25.2%) 33.9% (17.0–50.8%) F(1, 4) = 2.54, p = .19
3-Year NM SUD Initiators Persistent Users Comparison
20.8% (18.9–22.7%) 35.5% (32.2–38.7%) F(1, 35) = 67.16, p < .0001
16.6% (14.0–19.1%) 21.5% (18.0–25.0%) F(1, 28) = 5.11, p = .032
33.1% (31.6–34.6%) 54.0% (45.5–62.5%) F(1, 5) = 34.91, p = .002
Weekly NM Use at Maximum Quitters Persistent Users Comparison
44.2% (41.5–46.8%) 60.1% (55.4–64.9%) F(1, 23) = 38.95, p < .0001
37.0% (34.4–39.7%) 62.7% (59.4–66.0%) F(1, 26) = 172.47, p < .0001
56.6% (54.0–59.1%) 81.1% (73.6–88.7%) F(1, 7) = 93.84, p < .0001
Lifetime NM SUD Initiators (Ini) Quitters (Quit) Persistent Users (PU)
20.8% (18.8–22.7%) 26.2% (22.7–29.7%) 53.8% (49.8–57.8%)
16.6% (14.0–19.1%) 19.7% (17.4–22.0%) 40.2% (35.8–44.7%)
33.1% (29.4–36.8%) 43.1% (38.5–47.7%) 68.4% (62.4–74.3%)
Comparison Ini vs. Quit Quit vs. PU Ini vs. PU
F(2,90) = 117.20, p < .0001 F(1, 39) = 8.68, p = .005 F(1, 26) = 114.86, p < .0001 F(1, 35) = 269.34, p < .0001
F(2, 88) = 70.15, p < .0001 F(1, 39) = 3.19, p = .08 F(1, 27) = 153.49, p < .0001 F(1, 28) = 128.46, p < .0001
F(2, 28) = 75.52, p < .0001 F(1, 12) = 21.70, p = .0006 F(1, 7) = 154.01, p < .0001 F(1, 5) = 391.02, p < .0001
Notes: All comparisons used Rao-Scott chi square tests, after controlling for age, gender, race/ethnicity, marital status, education, household income, employment/student status, region of participant residence and past year DSM-IV Axis I and lifetime Axis II psychopathology; PY = Past Year NM = nonmedical; 3-year designates the period between waves 1 and 2 of the NESARC; SUD = substance use disorder.
(Oliva et al., 2012; Tucker et al., 2006), studies have found superior mental health in experimenters, as compared to never users. That said, between 37.0% (tranquilizer/sedative) and 56.6% (stimulant) of quitters engaged in weekly NUPM when using most frequently. Thus, while some proportion of quitters may have been infrequent users who experimented and quickly ceased use, at least one-third were more serious users, with at least 20% qualifying for an NUPMrelated SUD diagnosis. Furthermore, use of tranquilizer/sedative or stimulant medication may be to self-medicate untreated psychiatric conditions (e.g., anxiety disorders and ADHD, respectively), suggesting the potential for somewhat poorer ongoing mental health in some quitters. Perhaps a more likely hypothesis is that NUPM status moves in tandem with changes in health-related quality of life. Cessation may be a “virtuous cycle,” with improvements in health-related quality of life promoting continued cessation and continued cessation leading to further improvements in functioning. Initiation may be a “vicious cycle,” with NUPM and declines in health-related quality of life amplifying each other in a feedback loop. Indeed, NUPM can lead to increases in psychiatric problems (Schepis and Hakes, 2011, 2013), and psychiatric problems may lead to NUPM (Martins et al., 2012, 2009). NUPM is also longitudinally associated with physical health problems (Katz et al., 2013). Furthermore, healthrelated quality of life appears to improve in those who cease heavier use of other drugs (Best et al., 2013; Ugochukwu et al., 2013). Thus, quitters may cease NUPM because of improvements in mental and physical health-related quality of life, while initiators begin NUPM in response to worsening of mental and physical functioning. Conversely, NUPM cessation may improve and initiation may lead to declines in functioning. Unfortunately, it is impossible to offer more than conjecture about how NUPM and health-related quality of life interact, given the nature of the NESARC data. The largely equivalent change-based outcomes of persistent users and initiators were unexpected, as persisting substance use trajectories are associated with the poorest outcomes (Caldeira
et al., 2012; Schulenberg et al., 1996). Many persisting users may use at lower frequencies and amounts, suggesting a less problematic course, but the relatively high rates of weekly NUPM in the period of maximum use and NUPM-related SUD in persistent users argue against this. Furthermore, persistent users had the highest proportion of high school dropouts and individuals with household incomes under $40,000. The somewhat lower socioeconomic status of persistent users would be expected to negatively impact health (Adler and Rehkopf, 2008). Overall, the data tended to indicate that persistent users had notable longitudinal declines in health-related quality of life that seemed, at least in part, to be related to ongoing NUPM. 4.2. Differences based on medication class Generally speaking, the most robust significant differences in SF-12 change scores between NUPM subgroups were found for opioids and tranquilizer/sedatives. In contrast, significant differences in SF-12 change scores between stimulant subgroups were relatively few. While this may be because of lower base rates of stimulant NUPM and reduced power for analyses, it also may signal that stimulant NUPM is subject to quite different processes and produces different consequences than either opioids or tranquilizer/sedative NUPM. Important subgroup differences in SF-12 change score did appear, though, between opioid and tranquilizer/sedative nonmedical users. The subgroup-based changes in MCS and mental health-related quality of life subscales were very similar across opioids and tranquilizer/sedatives, with never users and quitters evidencing the best relative longitudinal outcomes and initiators and persistent users appearing to have greater declines in mental health-related quality of life. As with mental health-related quality of life, never users and quitters of opioids had the best relative changes in physical health-related quality of life, and initiators and persistent users suffered the greatest declines over time.
166
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
In contrast, subgroup differences in changes in physical healthrelated quality of life were minimal for tranquilizer/sedatives. Never users displayed marginally superior longitudinal profiles, and persistent users had surprising improvements in PCS scores between NESARC waves, as compared to quitters. Thus, while mental health-related quality of life and opioid or tranquilizer/sedative NUPM appear subject to similar interactions and processes, important differences in the relationship between physical health-related quality of life and opioid or tranquilizer/sedative NUPM appear to exist. This is an important avenue for future research, particularly in assessing potential moderators and mediators in these relationships. 4.3. Limitations Four limitations of this work should be noted. First, individuals were placed into the NUPM use groups based on any lifetime NUPM before wave 1 (initiators and persistent users) or any past year NUPM to wave 1 (quitters). Initiators and persistent users also had to have only at least one episode of NUPM between waves. This grouped infrequent nonmedical users with frequent and/or heavy nonmedical users, obscuring important differences based on NUPM frequency. Second, retrospective bias could have led to erroneous reports of past NUPM frequency and NUPM-related symptoms. Third, the NESARC assesses lifetime diagnosis of seven PDs at wave 1 and 3 at wave 2; assessment at two different timepoints could have influenced whether a respondent endorsed criteria for a PD, and lifetime diagnosis assessment could have led to some individuals being classified as having a PD despite having minimal symptoms at the time of assessment. Finally, although the response rate of the NESARC is excellent and statistical methods corrected for non-response, bias could have resulted from selective drop-out. Such bias should be minimal, though, if it exists (Kristman et al., 2004). 4.4. Conclusions These data underscore the importance of preventing NUPM initiation and encouraging cessation. Initiation was generally associated with poorest longitudinal outcomes for health-related quality of life, while cessation (or continued avoidance of NUPM) was associated with notably better changes in many aspects of health-related quality of life. Future investigations should followup on these results by examining potential moderators and mediators of the relationship between NUPM status and healthrelated quality of life, including frequency and/or severity of NUPM, source of medication for NUPM, motives for NUPM, medical conditions and psychopathology. Clinically speaking, this work highlights the importance of public health interventions to promote NUPM cessation and prevent initiation. Implementation of programs such as prescription monitoring databases (Gugelmann et al., 2012), clinician- or internet-based education programs on proper medication use, storage and disposal (McCauley et al., 2013) and universal substance use prevention programs (Spoth et al., 2013, 2008) could help prevent many of the adverse outcomes identified in this work among initiators and persistent users, reducing the significant morbidity, mortality and cost associated with NUPM. Role of the funding sources The NESARC is funded by the National Institute on Alcoholism and Alcohol Abuse. NIAAA had no further role in study design, the collection, analysis or interpretation of data, the writing of the report, or the decision to submit the paper for publication. Jahn Hakes is employed by the US Census Bureau. Any views expressed
on statistical and technical issues are those of the author(s) and not necessarily those of the U.S. Census Bureau. Contributors Ty Schepis was the primary writer of the manuscript. Jahn Hakes conducted all statistical analyses. Both authors participated equally in devising the research plan, selecting statistical analyses and in editing the manuscript. Both authors have read and approve the final version of the manuscript. Conflicts of interest The authors note no conflicts of interest. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.drugalcdep.2014.06.009. References Adler, N.E., Rehkopf, D.H., 2008. U.S. disparities in health: descriptions, causes, and mechanisms. Annu. Rev. Public Health 29, 235–252. American Psychiatric Association, 2000. Diagnostic And Statistical Manual Of Mental Disorders: DSM-IV-TR. American Psychiatric Association, Washington, DC. Best, D., Savic, M., Beckwith, M., Honor, S., Karpusheff, J., Lubman, D.I., 2013. The role of abstinence and activity in the quality of life of drug users engaged in treatment. J. Subst. Abuse Treat. 45, 273–279. Boyd, C.J., McCabe, S.E., Cranford, J.A., Young, A., 2006a. Adolescents’ motivations to abuse prescription medications. Pediatrics 118, 2472–2480. Boyd, C.J., McCabe, S.E., Teter, C.J., 2006b. Medical and nonmedical use of prescription pain medication by youth in a Detroit-area public school district. Drug Alcohol Depend. 81, 37–45. Caldeira, K.M., O’Grady, K.E., Vincent, K.B., Arria, A.M., 2012. Marijuana use trajectories during the post-college transition: health outcomes in young adulthood. Drug Alcohol Depend. 125, 267–275. Centers for Disease Control and Prevention, 2012. CDC Grand Rounds: Prescription Drug OVERDOSES—A U.S. Epidemic. MMWR 61, 10–13. Gandek, B., Ware, J.E., Aaronson, N.K., Apolone, G., Bjorner, J.B., Brazier, J.E., Bullinger, M., Kaasa, S., Leplege, A., Prieto, L., Sullivan, M., 1998. Cross-validation of item selection and scoring for the SF-12 Health Survey in nine countries: results from the IQOLA Project. International Quality of Life Assessment. J. Clin. Epidemiol. 51, 1171–1178. Garnier, L.M., Arria, A.M., Caldeira, K.M., Vincent, K.B., O’Grady, K.E., Wish, E.D., 2009. Nonmedical prescription analgesic use and concurrent alcohol consumption among college students. Am. J. Drug Alcohol Abuse 35, 334–338. Gonzalez-Saiz, F., Rojas, O.L., Castillo, I.I., 2009. Measuring the impact of psychoactive substance on health-related quality of life: an update. Curr. Drug Abuse Rev. 2, 5–10. Goodwin, R.D., Hasin, D.S., 2002. Sedative use and misuse in the United States. Addiction 97, 555–562. Grant, B.F., Dawson, D.A., Stinson, F.S., Chou, P.S., Kay, W., Pickering, R., 2003a. The Alcohol Use Disorder and Associated Disabilities Interview Schedule-IV (AUDADIS-IV): reliability of alcohol consumption, tobacco use, family history of depression and psychiatric diagnostic modules in a general population sample. Drug Alcohol Depend. 71, 7–16. Grant, B.F., Harford, T.C., Dawson, D.A., Chou, P.S., Pickering, R.P., 1995. The Alcohol Use Disorder And Associated Disabilities Interview Schedule (AUDADIS): reliability of alcohol and drug modules in a general population sample. Drug Alcohol Depend. 39, 37–44. Grant, B.F., Kaplan, K., Shepard, J., Moore, T., 2003b. Source And Accuracy Statement For Wave 1 Of The 2001-2002 National Epidemiologic Survey On Alcohol and Related Conditions. National Institute on Alcohol Abuse and Alcoholism, Bethesda, MD. Grant, B.F., Kaplan, K.D., 2005. Source and Accuracy Statement for the Wave 2 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). National Institute on Alcohol Abuse and Alcoholism, Rockville, MD. Gugelmann, H., Perrone, J., Nelson, L., 2012. Windmills and pill mills: can PDMPs tilt the prescription drug epidemic? J. Med. Toxicol. 8, 378–386. Havens, J.R., Young, A.M., Havens, C.E., 2011. Nonmedical prescription drug use in a nationally representative sample of adolescents: evidence of greater use among rural adolescents. Arch. Pediatr. Adolesc. Med. 165, 250–255. Hertz, J.A., Knight, J.R., 2006. Prescription drug misuse: a growing national problem. Adolesc. Med. Clin. 17, 751–769. Katz, C., El-Gabalawy, R., Keyes, K.M., Martins, S.S., Sareen, J., 2013. Risk factors for incident nonmedical prescription opioid use and abuse and dependence: results
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167 from a longitudinal nationally representative sample. Drug Alcohol Depend. 132, 107–113. Kim, H.M., Smith, E.G., Ganoczy, D., Walters, H., Stano, C.M., Ilgen, M.A., Bohnert, A.S., Valenstein, M., 2012. Predictors of suicide in patient charts among patients with depression in the Veterans Health Administration health system: importance of prescription drug and alcohol abuse. J. Clin. Psychiatry 73, e1269–e1275. Kristman, V., Manno, M., Cote, P., 2004. Loss to follow-up in cohort studies: how much is too much? Eur. J. Epidemiol. 19, 751–760. Kuramoto, S.J., Chilcoat, H.D., Ko, J., Martins, S.S., 2012. Suicidal ideation and suicide attempt across stages of nonmedical prescription opioid use and presence of prescription opioid disorders among U.S. adults. J. Stud. Alcohol Drugs 73, 178–184. Manchikanti, L., Singh, A., 2008. Therapeutic opioids: a ten-year perspective on the complexities and complications of the escalating use, abuse, and nonmedical use of opioids. Pain Phys. 11, S63–S88. Martins, S.S., Fenton, M.C., Keyes, K.M., Blanco, C., Zhu, H., Storr, C.L., 2012. Mood and anxiety disorders and their association with non-medical prescription opioid use and prescription opioid-use disorder: longitudinal evidence from the National Epidemiologic Study on Alcohol and Related Conditions. Psychol. Med. 42, 1261–1272. Martins, S.S., Keyes, K.M., Storr, C.L., Zhu, H., Chilcoat, H.D., 2009. Pathways between nonmedical opioid use/dependence and psychiatric disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Drug Alcohol Depend. 103, 16–24. Martins, S.S., Keyes, K.M., Storr, C.L., Zhu, H., Grucza, R.A., 2010. Birth-cohort trends in lifetime and past-year prescription opioid-use disorder resulting from nonmedical use: results from two national surveys. J. Stud. Alcohol Drugs 71, 480–487. McCabe, S.E., Knight, J.R., Teter, C.J., Wechsler, H., 2005. Non-medical use of prescription stimulants among US college students: prevalence and correlates from a national survey. Addiction 100, 96–106. McCabe, S.E., West, B.T., Wechsler, H., 2007. Trends and college-level characteristics associated with the non-medical use of prescription drugs among US college students from 1993 to 2001. Addiction 102, 455–465. McCauley, J.L., Back, S.E., Brady, K.T., 2013. Pilot of a brief, web-based educational intervention targeting safe storage and disposal of prescription opioids. Addict. Behav. 38, 2230–2235. Oliva, E.M., Keyes, M., Iacono, W.G., McGue, M., 2012. Adolescent substance use groups: antecedent and concurrent personality differences in a longitudinal study. J. Pers. 80, 769–793. Rao, J.N.K., Scott, A.J., 1984. On chi-squared tests for multi-way tables with cell proportions estimated from survey data. Ann. Stat. 12, 46–60. Schepis, T.S., Hakes, J.K., 2011. Nonmedical prescription use increases the risk for the onset and recurrence of psychopathology: results from the National
167
Epidemiological Survey on Alcohol and Related Conditions. Addiction 106, 2146–2155. Schepis, T.S., Hakes, J.K., 2013. Dose-related effects for the precipitation of psychopathology by opioid or tranquilizer/sedative nonmedical prescription use: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J. Addict. Med. 7, 39–44. Schepis, T.S., Krishnan-Sarin, S., 2008. Characterizing adolescent prescription misusers: a population-based study. J. Am. Acad. Child Adolesc. Psychiatry 47, 745–754. Schulenberg, J., O’Malley, P.M., Bachman, J.G., Wadsworth, K.N., Johnston, L.D., 1996. Getting drunk and growing up: trajectories of frequent binge drinking during the transition to young adulthood. J. Stud. Alcohol 57, 289–304. Shedler, J., Block, J., 1990. Adolescent drug use and psychological health. A longitudinal inquiry. Am. Psychol. 45, 612–630. Simoni-Wastila, L., Strickler, G., 2004. Risk factors associated with problem use of prescription drugs. Am. J. Public Health 94, 266–268. Spoth, R., Trudeau, L., Shin, C., Ralston, E., Redmond, C., Greenberg, M., Feinberg, M., 2013. Longitudinal effects of universal preventive intervention on prescription drug misuse: three randomized controlled trials with late adolescents and young adults. Am. J. Public Health 103, 665–672. Spoth, R., Trudeau, L., Shin, C., Redmond, C., 2008. Long-term effects of universal preventive interventions on prescription drug misuse. Addiction 103, 1160–1168. Substance Abuse and Mental Health Services Administration, 2010. Drug Abuse Warning Network, 2007: Area Profiles of Drug-Related Mortality. Office of Applied Studies, HHS Publication No. SMA 09-4407, DAWN Series D-31, Rockville, MD. Substance Abuse and Mental Health Services Administration, 2012. Results from the 2011 National Survey on Drug Use and Health: Summary of National Findings. NSDUH Series H-44, HHS Publication No. (SMA) 12-4713. Substance Abuse and Mental Health Services Administration, Rockville, MD. Substance Abuse and Mental Health Services Administration; Office of Applied Studies, 2009. Treatment Episode Data Set (TEDS) Highlights - 2007 National Admissions to Substance Abuse Treatment Services. OAS Series #S-45, HHS Publication No. (SMA) 09-4360, Rockville, MD. Tucker, J.S., Ellickson, P.L., Collins, R.L., Klein, D.J., 2006. Are drug experimenters better adjusted than abstainers and users? A longitudinal study of adolescent marijuana use. J. Adolesc. Health 39, 488–494. Ugochukwu, C., Bagot, K.S., Delaloye, S., Pi, S., Vien, L., Garvey, T., Bolotaulo, N.I., Kumar, N., Ishak, W.W., 2013. The importance of quality of life in patients with alcohol abuse and dependence. Harv. Rev. Psychiatry 21, 1–17. Ware Jr., J., Kosinski, M., Keller, S.D., 1996. A 12-Item short-form health survey: construction of scales and preliminary tests of reliability and validity. Med. Care 34, 220–233.
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Full length article
The association between nonmedical use of prescription medication status and change in health-related quality of life: Results from a Nationally Representative Survey Ty S. Schepis a,∗ , Jahn K. Hakes b a b
Department of Psychology, Texas State University, 601 University Drive, San Marcos, TX 78666, United States Center for Administrative Records Research and Applications, U.S. Census Bureau, Suitland, MD, United States
a r t i c l e
i n f o
Article history: Received 5 March 2014 Received in revised form 4 June 2014 Accepted 7 June 2014 Available online 19 June 2014 Keywords: Nonmedical prescription use Prescription misuse Opioid Tranquilizer Stimulant Sedative SF-12 Health-related quality of life
a b s t r a c t Background: Nonmedical use of prescription medication (NUPM) is associated cross-sectionally with a host of medical and psychosocial consequences. Few studies, however, have examined longitudinal outcomes based on NUPM indicators. This study aimed to address this gap by examining change in health-related quality of life as a function of NUPM status. Methods: Data are from waves 1 and 2 of the National Epidemiological Survey on Alcoholism and Related Conditions (NESARC) a household-based, nationally-representative survey of the US population. 34,653 participants who completed both NESARC waves were included in analyses. The primary outcome measure was the 12-item Short Form Health Survey (SF-12), with history of NUPM of opioids, tranquilizer/sedatives and stimulants (examined separately) at wave 1 and any NUPM between waves 1 and 2 used to group participants. Sociodemographic characteristics were used as control variables. Results: Across medication classes, results indicated that individuals who initiated NUPM between waves (initiators) had greater declines or smaller increases on many SF-12 scales, when compared to other groups. Individuals with a history of NUPM at wave 1 but no use between waves (quitters) and never users generally had the best outcomes in terms of change in SF-12 scales, with quitters making larger gains (or smaller losses) in mental health-related quality of life. Persistent users were generally intermediate between initiators and quitters or never users. Conclusions: These data reinforce the importance of preventing NUPM initiation and of promoting NUPM cessation, highlighting the need for greater use of NUPM-related public health interventions. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The nonmedical use of prescription medications (NUPM) in the United States has been described as a public health concern (Garnier et al., 2009; Schepis and Hakes, 2013) or even an epidemic (Centers for Disease Control and Prevention, [CDC], 2012), with rates of NUPM and NUPM-related substance use disorder (SUD) having increased notably over the past 20 years (Martins et al., 2010; McCabe et al., 2007). While rates have leveled off somewhat, rates of opioid NUPM trail only those of alcohol, tobacco and marijuana (Substance Abuse and Mental Health Services Administration [SAMHSA], 2012); when combined with NUPM from other classes of medication (e.g., tranquilizers, sedatives and stimulants), the
∗ Corresponding author. Tel.: +1 512 245 6805; fax: +1 512 245 3153. E-mail address: [email protected] (T.S. Schepis). http://dx.doi.org/10.1016/j.drugalcdep.2014.06.009 0376-8716/© 2014 Elsevier Ireland Ltd. All rights reserved.
total past year prevalence of NUPM is 5.6%, which is larger than the pooled past year prevalence rates (3.9%) of cocaine, methamphetamine, heroin or any hallucinogen use (SAMHSA, 2012). With increases in NUPM rates, rates of NUPM-related emergency department utilization, substance use treatment and overdose have increased dramatically (CDC, 2012; Manchikanti and Singh, 2008; SAMHSA, 2009, 2010). Moreover, substantial crosssectional research has associated NUPM with serious psychosocial and neurobehavioral consequences, including psychiatric illness (Goodwin and Hasin, 2002), suicidality (Kim et al., 2012; Kuramoto et al., 2012), other substance use (Boyd et al., 2006b; Schepis and Krishnan-Sarin, 2008), poorer academic performance (McCabe et al., 2005), unemployment (Simoni-Wastila and Strickler, 2004) and poorer self-reported health (Havens et al., 2011). Longitudinal research indicates that NUPM is associated with increased incidence and recurrence of psychiatric disorders (Martins et al., 2012; Schepis and Hakes, 2011) and that psychiatric disorders may
162
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
increase the likelihood of NUPM and NUPM-related SUD (Martins et al., 2012). Longitudinal data also indicate that certain physical conditions increase the risk of NUPM-related SUD at a 3-year follow-up (Katz et al., 2013). Despite these initial efforts, more longitudinal work is needed to examine the processes that promote NUPM and the consequences of NUPM. To illustrate, no research has examined how NUPM course longitudinally affects health-related quality of life. Health-related quality of life is the individual’s perception of his or her physical, mental and social functioning and how such functioning affects his or her well-being and daily activities (Gonzalez-Saiz et al., 2009). The evidence across substance users, with the greatest number of investigations in alcohol and opioid users, indicates that substance users have poorer health-related quality of life than non-users, with the worst health-related quality of life in those with more frequent or severe use (Gonzalez-Saiz et al., 2009; Ugochukwu et al., 2013). In turn, abstinence from substance use is associated with improvements in health-related quality of life (Best et al., 2013; Ugochukwu et al., 2013). As noted above, NUPM increases the risk of a psychiatric diagnosis over a 3-year follow-up (Martins et al., 2012; Schepis and Hakes, 2011), but no published work has investigated the longitudinal impact of NUPM on a more continuous measure of mental health-related quality of life. While use of categorical diagnoses provides important data about NUPM-related mental health outcomes, diagnostic categorization may miss larger trends toward declining mental health-related quality of life in nonmedical users, especially if such declines are not severe enough in many to warrant a diagnosis. Examination of the longitudinal change in self-reported physical and mental health-related quality of life could provide important data on the consequences of NUPM. 1.1. Aims and hypotheses This work aims to accomplish this goal through the use of two waves of data from the National Epidemiological Survey on Alcohol and Related Conditions (NESARC). Both waves of the NESARC contain comprehensive assessments of NUPM using opioids, tranquilizers, stimulants and sedatives and the 12-item Short Form Health Survey (SF-12), a validated measure of physical and mental health-related quality of life (Ware et al., 1996). The primary aim was to evaluate changes in SF-12 scores over the 3-year period between NESARC waves by NUPM course: never users (denied any NUPM at both waves), initiators (denied lifetime NUPM at wave 1, but endorsed NUPM between waves), quitters (endorsed past year NUPM at wave 1, but denied NUPM between waves) and persistent users (endorsed lifetime NUPM at wave 1 and between waves). Analyses were performed separately by medication group, with tranquilizers and sedatives combined in a single group; comparisons examined differences in SF-12 scale and subscale change scores between participant subgroups. Across medication class, we hypothesized that never users would have the largest gains or smallest declines in SF-12 scores over the follow-up, followed by quitters and initiators. We also hypothesized that persistent users would have the worst outcomes, as signified by the largest decreases or smallest gains on the SF-12. 2. Methods The NESARC is a longitudinal survey funded by the National Institute on Alcohol Abuse and Alcoholism (NIAAA), with two available data waves (wave 1: 2001–2002; wave 2: 2004–2005). The NESARC aims to provide a representative sample of the US adult population for analyses of substance use behaviors and correlates. The 2000–2001 Census Supplementary Survey structured the sampling, which is described elsewhere (Grant et al., 2003b; Grant and Kaplan, 2005). The NESARC includes weighting to produce nationally representative data, and to adjust for survey design, young adult oversampling, and non-response at both wave 1 and wave 2 (among wave 1 participants). The US Census Bureau and the US Office
of Management and Budget approved the NESARC protocol, and the Texas State University IRB approved this investigation. At wave 1, 43,093 individuals participated with 39,959 eligible to participate in wave 2. Of those eligible, 34,653 consented to participate in wave 2. The wave 1 response rate was 81.2%, and the wave 2 response rate was 86.7%, with a total response rate of 70.2% (Grant et al., 2003b; Grant and Kaplan, 2005). The weighted NESARC sample is 52% female, 71% Caucasian, 12% Hispanic/Latino and 11% AfricanAmerican; 13% was under 25 years of age.
2.1. Measures 2.1.1. Lifetime nonmedical use of prescription medication (NUPM). In the NESARC, NUPM is defined as prescription use “without a prescription, in greater amounts, more often, or longer than prescribed, or for a reason other than a doctor said you should use them.” Lifetime NUPM and NUPM between waves 1 and 2 were assessed for four classes of medication: opioids, tranquilizers, stimulants and sedatives. Due to low base rates for sedative NUPM, similar pharmacological properties and motives for sedative and tranquilizer NUPM (Boyd et al., 2006a; Hertz and Knight, 2006), sedative and tranquilizer NUPM were pooled.
2.1.2. Frequency of NUPM. For participants who endorsed lifetime NUPM from a given medication class, follow-up questions are asked at wave 1 to determine lifetime maximum use frequency and at wave 2 to assess frequency of NUPM within the past year.
2.1.3. Substance Use Disorder (SUD) from NUPM. SUD diagnosis was obtained through the NIAAA Alcohol Use Disorder and Associated Disabilities Interview Schedule–DSM-IV Edition (AUDADIS-IV; Grant et al., 2003a, 1995). The AUDADIS-IV is a fully structured diagnostic interview that assesses symptoms of many DSMIV (American Psychiatric Association, 2000) disorders. The AUDADIS-IV appears to have good reliability and validity in assessing SUD (Grant et al., 2003a, 1995). At wave 1, lifetime SUD from NUPM was captured, and at wave 2, SUD from NUPM during the follow-up period was assessed.
2.1.4. Axis I diagnosis. Here, Axis I psychiatric outcomes assessed by the AUDADISIV include two depressive disorders (MDD or dysthymia), two bipolar disorders (bipolar I or II), six anxiety disorders (panic disorder with and without agoraphobia, agoraphobia without panic, social phobia, specific phobia or generalized anxiety disorder) and pathological gambling. Disorders were assessed at wave 1, were for past year diagnosis and were included in models as a dichotomous control variable.
2.1.5. Axis II diagnosis. Seven lifetime Axis II personality disorder (PD) diagnoses assessed at wave 1 and three PDs assessed at wave 2 were included in models as a dichotomous control variable. PDs assessed at wave 1 were antisocial, avoidant, dependent, obsessive-compulsive, paranoid, schizoid, and histrionic PD, with borderline, schizotypal and narcissistic PDs assessed at wave 2. The PDs assessed at wave 1 were only assessed at that timepoint, and the PDs assessed at wave 2 were only assessed at that point; in both cases only lifetime PD diagnosis was assessed.
2.1.6. 12-item Short-Form Health Survey, version 2 (SF-12). The SF-12 is a shorter version of the 36-item Short-Form Health Survey, with comparable reliability and validity indicators (Gandek et al., 1998; Ware et al., 1996). The SF-12 contains questions about physical and mental health-related quality of life, answered on a 5-point Likert-type scale. The SF-12 produces two summary scales, the physical component and mental component scales (PCS and MCS, respectively) and eight subscales. The physical functioning, role-physical, bodily pain and general health subscales compose the larger PCS scale, and the vitality, social functioning, role-emotional and mental health subscales compose the larger MCS scale. Sociodemographic variables assessed were age, gender, race/ethnicity, marital status, education level, household income, employment/full-time student status, and region of participant residence.
2.2. Subsamples Analyses split participants into 4 groups for each class of medication, based on NUPM status at wave 1 and status between waves. Participants who denied NUPM at both waves were grouped as never users, while those who denied lifetime NUPM at wave 1 but endorsed NUPM between waves were described as initiatiors. Participants who endorsed both lifetime NUPM at wave 1 and NUPM between waves were grouped as persistent users. The only exception to the lifetime NUPM classification was in quitters, who were classified as those endorsing NUPM in the year prior to wave 1 but no NUPM between waves. Quitters were so classified to isolate the effects of NUPM cessation, as using “lifetime quitters” who engaged in lifetime NUPM with no NUPM between waves would have included many individuals who ceased NUPM many years prior. Participants with lifetime NUPM but no past year NUPM and no NUPM between NESARC waves were excluded from analyses.
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
163
Table 1 Means of SF-12 score change from wave 1 to wave 2 by opioid nonmedical use group.
Physical Component Physical Functioning Role Physical Bodily Pain General Health Mental Component Vitality Social Functioning Role Emotional Mental Health
Never Users (1) (n = 32419; 95.8%)
Initiators (2) (n = 735; 2.2%)
Quitters (3) (n = 412; 2.2%)
Persistent Users (4) (n = 262; 0.8%)
Comparisons between groups, by group number
−0.64 (−0.70, −0.59) −0.90 (−0.96, −0.84) −1.47 (−1.54, −1.40) 0.43 (0.36, 0.50) −1.10 (−1.18, −1.01) −1.18 (−1.27, −1.09) −1.78 (−1.89, −1.67) −0.51 (−0.60, −0.42) −1.89 (−1.99, −1.79) −0.35 (−0.43, −0.26)
−1.43 (−1.72, −1.15) −1.41 (−1.67, −1.14) −2.45 (−2.84, −2.07) −1.10 (−1.74, −0.45) −1.51 (−1.91, −1.11) −2.19 (−2.74, −1.63) −2.41 (−2.88, −1.95) −2.10 (−2.82, −1.38) −2.91 (−3.40, −2.43) −1.14 (−1.70, −0.58)
0.84 (0.59, 1.09) −0.48 (−0.89, −0.08) 0.59 (0.13, 1.05) 2.06 (1.52, 2.59) 0.93 (0.64, 1.22) −0.11 (−0.53, 0.31) −0.60 (−0.93, −0.27) 0.55 (−0.05, 1.16) −0.98 (−1.69, −0.27) 0.97 (0.45, 1.48)
−1.38 (−1.82, −0.94) −1.77 (−2.09, −1.45) −3.28 (−3.81, −2.74) −0.45 (−1.20, 0.30) −0.82 (−1.68, 0.03) −2.50 (−3.41, −1.60) −3.36 (−3.98, −2.74) -1.63 (−2.49, −0.77) −3.12 (−4.19, −2.04) −1.63 (−2.63, −0.63)
3 > all; 1 > 2 1 > 2, 4 3 > all; 1 > 2, 4 3 > 2, 4; 1 > 2 3 > all 1, 3 > 2, 4 1, 3 > 4 1, 3 > 2, 4 1 > 2, 4 3 > 2, 4; 1 > 4
Notes: Means of change from wave 1 to wave 2 are listed with 95% confidence intervals in parentheses; n.s.d = no significant differences; the “Comparisons” column uses multivariate regression to compare groups, after controlling for age, gender, race/ethnicity, marital status, education, household income, employment/student status, region of participant residence and past year DSM-IV Axis I and lifetime Axis II psychopathology, with ˛ = 0.05.
2.3. Data analysis SF-12 mean scores for each medication class and each subgroup of participants were calculated, along with frequencies/percentages of sociodemographic characteristics for each subgroup. Primary analyses used design-based linear regression, comparing change in SF-12 score for each component scale and subscale between members of each subgroup. Analyses were run seperately by class of medication, and analyses controlled for the 8 sociodemographic variables listed above, past year (from wave 1) Axis I psychiatric diagnosis and lifetime Axis II diagnosis. Followup analyses examined three nonmedical use characteristics between initiators and persistent users: (1) past year nonmedical use, (2) past year weekly nonmedical use, and (3) SUD from nonmedical use in the period between wave 1 and wave 2. Followup analyses also examined potential differences in weekly nonmedical use at the lifetime maximum use level between quitters and persistent users. Finally, analyses compared initiators, quitters and persisitent users on rates of lifetime NUPM-related SUD. Follow-up analyses used Rao-Scott chi-square (Rao and Scott, 1984) to compare prevalence rates seperately by class of medication. All data were weighted, clustered on primary sampling units, and stratified appropriately. We included models only if they evidenced adequate fit. All analyses were performed in SAS, version 9.2 (Cary, NC).
3. Results 3.1. Subsample prevalence and sociodemographics The largest participant group was never users (opioids: 95.8%; tranquilizer/sedatives: 96.0%; stimulants: 99.1%), followed by initiators (opioids: 2.2%; tranquilizer/sedatives: 1.9%; stimulants: 0.4%) and quitters (opioids: 1.2%; tranquilizer/sedatives: 1.3%; stimulants: 0.4%). Persistent users were the smallest group (opioids: 0.8%; tranquilizer/sedatives: 0.8%; stimulants: 0.2%). Sociodemographic and NUPM characteristics for each group are included as Supplementary Material.1 Across medication classes, a higher proportion of females and non-Caucasian participants were never users, with the second highest proportion in the initiators group. Persistent users had the highest proportion of male and Caucasian participants. Otherwise, patterns were less consistent. Never users tended to be older and more likely to be married, with initiators and persistent users having higher proportions in the 18–25 cohort and the never married group. Persistent users and quitters were more likely to have household incomes below $40,000, though this did not clearly hold for tranquilizer/sedative users.
Table 1. For change in PCS score, quitters’ scores improved while all other subgroups’ scores declined, and never users had a smaller decline over the follow-up than opioid NUPM initiators. All PCS subscales evidenced significant subgroup differences: one, never users had smaller declines in physical functioning than initiators and persistent users; two, quitters evidenced improvements on the role physical subscale while all other subgroups declined, and never users had smaller declines than initiators or persistent users; three, quitters had better change outcomes on the bodily pain subscale than initiators or persistent users, and never users had superior follow-up outcomes than initiators; finally, quitters had improvements in general health between waves while all other subgroups evidenced declines. For change in MCS score and the social functioning subscale, opioid NUPM quitters and never users had significantly more adaptive changes in functioning than initiators or persistent users. Both quitters and never users had better longitudinal outcomes on the vitality and mental health subscales than persistent users; quitters were also superior to initiators in change in the mental health and social functioning subscales. Finally, never users had smaller declines between waves on the role emotional subscale than initiators or persistent users. 3.3. Change in SF-12 scores by tranquilizer/sedative nonmedical use group All means and 95% confidence intervals for SF-12 change scores by tranquilizer/sedative subgroup are listed in Table 2. The PCS and its subscales evidenced an inconsistent pattern of longitudinal results, with tranquilizer/sedative persistent users having significantly better change in PCS scores than quitters. Never users had smaller declines between waves in the physical functioning subscale than either initiators or quitters, and never users had smaller declines on the role-emotional subscale than initiators. No significant differences were found on the bodily pain or general health subscales. On the MCS, and the social functioning and the mental health subscales, never users and quitters had superior longitudinal outcomes than both initiators and persistent users. For the vitality and role-emotional subscales, the never user and quitter subgroups evidenced smaller declines in change scores than did initiators.
3.2. Change in SF-12 scores by opioid nonmedical use group 3.4. Change in SF-12 scores by stimulant nonmedical use group All means and 95% confidence intervals for SF-12 component and subscale change scores by opioid use subgroup are listed in
1 Supplementary material can be found by accessing the online version of this paper at http://dx.doi.org and by entering doi:. . ..
All means and 95% confidence intervals for SF-12 change scores by stimulant subgroup are listed in Table 3. Comparison of the stimulant subgroups demonstrated many fewer significant differences in change scores than those seen in the opioid or tranquilizer/sedative comparisons. Quitters had superior longitudinal
164
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
Table 2 Means of SF-12 score change from wave 1 to wave 2 by tranquilizer/sedative nonmedical use group.
Physical Component Physical Functioning Role Physical Bodily Pain General Health Mental Component Vitality Social Functioning Role Emotional Mental Health
Never Users (1) (n = 32232; 96.0%)
Initiators (2) (n = 653; 1.9%)
Quitters (3) (n = 427; 1.3%)
Persistent Users (4) (n = 252; 0.8%)
Comparisons between groups, by group number
−0.65 (−0.71, −0.59) −0.89 (−0.95, −0.83) −1.47 (−1.54, −1.41) 0.40 (0.33, 0.48) −1.14 (−1.22, −1.05) −1.19 (−1.28, −1.11) −1.77 (−1.87, −1.66) −0.54 (−0.62, −0.45) −1.90 (−2.00, −1.80) −0.36 (−0.44, −0.27)
−0.79 (−1.06, −0.53) −1.63 (−1.91, −1.36) −2.44 (−2.81, −2.07) −0.42 (−0.85, 0.003) −1.67 (−1.96, −1.38) −3.92 (−4.48, −3.35) −3.58 (−4.03, −3.13) −2.67 (−3.22, −2.12) −3.70 (−4.22, −3.17) −3.09 (−3.61, −2.56)
−1.22 (−1.55, −0.89) −1.58 (−1.96, −1.21) −1.93 (−2.19, −1.66) 0.24 (−0.19, 0.66) −0.24 (−0.60, 0.11) 0.29 (−0.07, 0.64) −0.53 (−0.97, −0.09) 0.48 (0.14, 0.82) −1.63 (−1.98, −1.28) 0.83 (0.47, 1.18)
0.61 (0.25, 0.98) −0.74 (−1.04, −0.44) −1.26 (−1.79, −0.72) 2.10 (0.90, 3.29) 0.04 (−0.61, 0.70) −2.60 (−3.89, −1.32) −2.22 (−2.92, −1.52) −2.03 (−3.03, −1.03) −2.15 (−3.57, −0.74) −1.49 (−2.64, −0.34)
4>3 1 > 2, 3 1>2 n.s.d. n.s.d. 1, 3 > 2, 4 1, 3 > 2 1, 3 > 2, 4 1, 3 > 2 1, 3 > 2, 4
Notes: Means of change from wave 1 to wave 2 are listed with 95% confidence intervals in parentheses; n.s.d = no significant differences; the “Comparisons” column uses multivariate regression to compare groups, after controlling for age, gender, race/ethnicity, marital status, education, household income, employment/student status, region of participant residence and past year DSM-IV Axis I and lifetime Axis II psychopathology, with ˛ = 0.05.
outcomes than never users on the PCS, and the change scores of quitters were superior to all groups on the general health change score. Other physical health-related quality of life subscales (e.g., physical functioning, role physical, or bodily pain) did not evidence significant subgroup differences in change scores. Similarly, the MCS scale, role emotional and mental health subscales did not indicate significant changes in scores between stimulant subgroups. Only two mental health-related quality of life subscales displayed significant differences: both the vitality and social functioning subscales indicated that never users and quitters had superior longitudinal outcomes over initiators. 3.5. Comparison of NUPM characteristics between use groups In order to further characterize the three groups engaged in lifetime NUPM (i.e., initiators, quitters and persistent users), their nonmedical use characteristics were compared. Across medication class, persistent users had higher rates of past year (to wave 2) NUPM and NUPM-related SUD diagnosis during the follow-up period than initiators. Conversely, past year rates of weekly NUPM did not differ between initiators and persistent users for any medication class. Rates of weekly NUPM at the respondent’s maximum level of NUPM were significantly higher for quitters than persistent users across medication classes. Finally, across medication classes, rates of lifetime NUPM-related SUD were highest in persistent users, followed by quitters. Initiators had the lowest rates of lifetime NUPM-related SUD, and all three groups were significantly different with one exception: comparing initiators and quitters for tranquilizer/sedative-related SUD. Prevalence rates with 95% confidence intervals for these NUPM characteristics and the Wald F-test output are in Table 4.
4. Discussion This investigation produced both expected (e.g., the relatively adaptive longitudinal outcomes of never users) and unexpected results (e.g., the superior longitudinal health-related quality of life outcomes of quitters and the particularly poor follow-up outcomes seen in initiators) in light of our stated hypotheses. Across medication classes, never users generally had superior outcomes to initiators and persistent users, as expected. Contrary to our expectations, though, quitters often had roughly equivalent outcomes to never users across the SF-12 scale change scores, with superior outcomes to initiators and persistent users, particularly on mental health-related quality-of-life scales. These patterns were much less apparent for the stimulants, but they largely held for the opioid and tranquilizer/sedative subgroups. Along with the finding that NUPM cessation was beneficial, the other key finding of this work was that NUPM initiation was associated with notably greater declines in health-related quality of life, relative to never users and quitters.
4.1. Differences based on NUPM subgroup While we expected that never users would have better changes in SF-12 scores over the three-year follow-up period than any other group, we found that quitters were roughly equivalent on SF-12 change scores. For the PCS scale and the physical component subscales, the change in health-related quality of life for quitters was superior to even never users for opioid NUPM. The equivalent longitudinal mental component profile of quitters to never users may suggest that most of these users would fall into what is called an “experimenter” group, using infrequently before ceasing use. Starting with the work of Shedler and Block (1990), many, but not all
Table 3 Change from wave 1 to wave 2 in SF-12 scores by stimulant nonmedical use group.
Physical Component Physical Functioning Role Physical Bodily Pain General Health Mental Component Vitality Social Functioning Role Emotional Mental Health
Never Users (1) (n = 33031; 99.1%)
Initiators (2) (n = 125; 0.4%)
Quitters (3) (n = 123; 0.4%)
Persistent Users (4) (n = 51; 0.2%)
Comparisons between groups, by group number
−0.66 (−0.71, −0.60) −0.94 (−1.00, −0.88) −1.49 (−1.56, −1.43) 0.42 (0.35, 0.50) −1.12 (−1.21, −1.04) −1.21 (−1.30, −1.12) −1.77 (−1.88, −1.67) −0.54 (−0.62, −0.46) −1.94 (−2.03, −1.84) −0.38 (−0.47, −0.29)
−1.40 (−2.68, −0.12) −1.32 (−2.54, −0.10) −1.98 (−2.93, −1.03) −0.92 (−3.42, 1.58) −2.11 (−3.48, −0.73) −3.38 (−4.30, −2.45) −5.21 (−5.80, −4.61) −3.95 (−4.87, −3.03) −2.39 (−2.86, −1.93) −1.49 (−2.25, −0.74)
1.76 (0.23, 3.29) 0.42 (−0.03, 0.87) 0.30 (−0.77, 1.37) 2.49 (0.57, 4.41) 2.07 (1.76, 2.37) −0.86 (−2.58, 0.85) −0.79 (−1.86, 0.28) 0.67 (0.15, 1.19) −1.42 (−2.72, −0.13) 0.35 (−1.13, 1.83)
−0.52 (−1.79, 0.74) −2.78 (−4.37, −1.20) −1.65 (−2.51, −0.80) 2.94 (1.09, 4.79) −2.24 (−2.35, −2.13) −2.54 (−6.80, 1.71) −2.09 (−2.47, −1.70) −2.36 (−10.04, 5.32) −1.57 (−1.76, −1.37) −2.75 (−6.27, 0.77)
3>2 n.s.d. n.s.d. n.s.d. 3 > all n.s.d. 1, 3 > 2 1, 3 > 2 n.s.d. n.s.d.
Notes: Means of change from wave 1 to wave 2 are listed with 95% confidence intervals in parentheses; n.s.d = no significant differences; the “Comparisons” column uses multivariate regression to compare groups, after controlling for age, gender, race/ethnicity, marital status, education, household income, employment/student status, region of participant residence and past year DSM-IV Axis I and lifetime Axis II psychopathology, with ˛ = 0.05.
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
165
Table 4 Subgroup nonmedical use characteristics examined separately by medication class. Opioids
Tranquilizer/Sedative
Stimulant
PY NM Use Initiators Persistent Users Comparison
67.1% (65.1–69.2%) 75.2% (71.8–78.7%) F(1, 35) = 15.09, p = .0004
65.5% (63.9–67.2%) 75.1% (71.6–78.6%) F(1, 28) = 25.34, p < .0001
65.2% (62.8–67.6%) 79.9% (76.2–83.5%) F(1, 5) = 81.09, p = .0003
PY Weekly NM Use Initiators Persistent Users Comparison
25.8% (24.0–27.6%) 20.2% (14.6–25.8%) F(1, 22) = 3.88, p = .062
23.2% (21.3–25.2%) 20.8% (16.9–24.7%) F(1, 18) = 1.07, p = .31
23.6% (22.0–25.2%) 33.9% (17.0–50.8%) F(1, 4) = 2.54, p = .19
3-Year NM SUD Initiators Persistent Users Comparison
20.8% (18.9–22.7%) 35.5% (32.2–38.7%) F(1, 35) = 67.16, p < .0001
16.6% (14.0–19.1%) 21.5% (18.0–25.0%) F(1, 28) = 5.11, p = .032
33.1% (31.6–34.6%) 54.0% (45.5–62.5%) F(1, 5) = 34.91, p = .002
Weekly NM Use at Maximum Quitters Persistent Users Comparison
44.2% (41.5–46.8%) 60.1% (55.4–64.9%) F(1, 23) = 38.95, p < .0001
37.0% (34.4–39.7%) 62.7% (59.4–66.0%) F(1, 26) = 172.47, p < .0001
56.6% (54.0–59.1%) 81.1% (73.6–88.7%) F(1, 7) = 93.84, p < .0001
Lifetime NM SUD Initiators (Ini) Quitters (Quit) Persistent Users (PU)
20.8% (18.8–22.7%) 26.2% (22.7–29.7%) 53.8% (49.8–57.8%)
16.6% (14.0–19.1%) 19.7% (17.4–22.0%) 40.2% (35.8–44.7%)
33.1% (29.4–36.8%) 43.1% (38.5–47.7%) 68.4% (62.4–74.3%)
Comparison Ini vs. Quit Quit vs. PU Ini vs. PU
F(2,90) = 117.20, p < .0001 F(1, 39) = 8.68, p = .005 F(1, 26) = 114.86, p < .0001 F(1, 35) = 269.34, p < .0001
F(2, 88) = 70.15, p < .0001 F(1, 39) = 3.19, p = .08 F(1, 27) = 153.49, p < .0001 F(1, 28) = 128.46, p < .0001
F(2, 28) = 75.52, p < .0001 F(1, 12) = 21.70, p = .0006 F(1, 7) = 154.01, p < .0001 F(1, 5) = 391.02, p < .0001
Notes: All comparisons used Rao-Scott chi square tests, after controlling for age, gender, race/ethnicity, marital status, education, household income, employment/student status, region of participant residence and past year DSM-IV Axis I and lifetime Axis II psychopathology; PY = Past Year NM = nonmedical; 3-year designates the period between waves 1 and 2 of the NESARC; SUD = substance use disorder.
(Oliva et al., 2012; Tucker et al., 2006), studies have found superior mental health in experimenters, as compared to never users. That said, between 37.0% (tranquilizer/sedative) and 56.6% (stimulant) of quitters engaged in weekly NUPM when using most frequently. Thus, while some proportion of quitters may have been infrequent users who experimented and quickly ceased use, at least one-third were more serious users, with at least 20% qualifying for an NUPMrelated SUD diagnosis. Furthermore, use of tranquilizer/sedative or stimulant medication may be to self-medicate untreated psychiatric conditions (e.g., anxiety disorders and ADHD, respectively), suggesting the potential for somewhat poorer ongoing mental health in some quitters. Perhaps a more likely hypothesis is that NUPM status moves in tandem with changes in health-related quality of life. Cessation may be a “virtuous cycle,” with improvements in health-related quality of life promoting continued cessation and continued cessation leading to further improvements in functioning. Initiation may be a “vicious cycle,” with NUPM and declines in health-related quality of life amplifying each other in a feedback loop. Indeed, NUPM can lead to increases in psychiatric problems (Schepis and Hakes, 2011, 2013), and psychiatric problems may lead to NUPM (Martins et al., 2012, 2009). NUPM is also longitudinally associated with physical health problems (Katz et al., 2013). Furthermore, healthrelated quality of life appears to improve in those who cease heavier use of other drugs (Best et al., 2013; Ugochukwu et al., 2013). Thus, quitters may cease NUPM because of improvements in mental and physical health-related quality of life, while initiators begin NUPM in response to worsening of mental and physical functioning. Conversely, NUPM cessation may improve and initiation may lead to declines in functioning. Unfortunately, it is impossible to offer more than conjecture about how NUPM and health-related quality of life interact, given the nature of the NESARC data. The largely equivalent change-based outcomes of persistent users and initiators were unexpected, as persisting substance use trajectories are associated with the poorest outcomes (Caldeira
et al., 2012; Schulenberg et al., 1996). Many persisting users may use at lower frequencies and amounts, suggesting a less problematic course, but the relatively high rates of weekly NUPM in the period of maximum use and NUPM-related SUD in persistent users argue against this. Furthermore, persistent users had the highest proportion of high school dropouts and individuals with household incomes under $40,000. The somewhat lower socioeconomic status of persistent users would be expected to negatively impact health (Adler and Rehkopf, 2008). Overall, the data tended to indicate that persistent users had notable longitudinal declines in health-related quality of life that seemed, at least in part, to be related to ongoing NUPM. 4.2. Differences based on medication class Generally speaking, the most robust significant differences in SF-12 change scores between NUPM subgroups were found for opioids and tranquilizer/sedatives. In contrast, significant differences in SF-12 change scores between stimulant subgroups were relatively few. While this may be because of lower base rates of stimulant NUPM and reduced power for analyses, it also may signal that stimulant NUPM is subject to quite different processes and produces different consequences than either opioids or tranquilizer/sedative NUPM. Important subgroup differences in SF-12 change score did appear, though, between opioid and tranquilizer/sedative nonmedical users. The subgroup-based changes in MCS and mental health-related quality of life subscales were very similar across opioids and tranquilizer/sedatives, with never users and quitters evidencing the best relative longitudinal outcomes and initiators and persistent users appearing to have greater declines in mental health-related quality of life. As with mental health-related quality of life, never users and quitters of opioids had the best relative changes in physical health-related quality of life, and initiators and persistent users suffered the greatest declines over time.
166
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167
In contrast, subgroup differences in changes in physical healthrelated quality of life were minimal for tranquilizer/sedatives. Never users displayed marginally superior longitudinal profiles, and persistent users had surprising improvements in PCS scores between NESARC waves, as compared to quitters. Thus, while mental health-related quality of life and opioid or tranquilizer/sedative NUPM appear subject to similar interactions and processes, important differences in the relationship between physical health-related quality of life and opioid or tranquilizer/sedative NUPM appear to exist. This is an important avenue for future research, particularly in assessing potential moderators and mediators in these relationships. 4.3. Limitations Four limitations of this work should be noted. First, individuals were placed into the NUPM use groups based on any lifetime NUPM before wave 1 (initiators and persistent users) or any past year NUPM to wave 1 (quitters). Initiators and persistent users also had to have only at least one episode of NUPM between waves. This grouped infrequent nonmedical users with frequent and/or heavy nonmedical users, obscuring important differences based on NUPM frequency. Second, retrospective bias could have led to erroneous reports of past NUPM frequency and NUPM-related symptoms. Third, the NESARC assesses lifetime diagnosis of seven PDs at wave 1 and 3 at wave 2; assessment at two different timepoints could have influenced whether a respondent endorsed criteria for a PD, and lifetime diagnosis assessment could have led to some individuals being classified as having a PD despite having minimal symptoms at the time of assessment. Finally, although the response rate of the NESARC is excellent and statistical methods corrected for non-response, bias could have resulted from selective drop-out. Such bias should be minimal, though, if it exists (Kristman et al., 2004). 4.4. Conclusions These data underscore the importance of preventing NUPM initiation and encouraging cessation. Initiation was generally associated with poorest longitudinal outcomes for health-related quality of life, while cessation (or continued avoidance of NUPM) was associated with notably better changes in many aspects of health-related quality of life. Future investigations should followup on these results by examining potential moderators and mediators of the relationship between NUPM status and healthrelated quality of life, including frequency and/or severity of NUPM, source of medication for NUPM, motives for NUPM, medical conditions and psychopathology. Clinically speaking, this work highlights the importance of public health interventions to promote NUPM cessation and prevent initiation. Implementation of programs such as prescription monitoring databases (Gugelmann et al., 2012), clinician- or internet-based education programs on proper medication use, storage and disposal (McCauley et al., 2013) and universal substance use prevention programs (Spoth et al., 2013, 2008) could help prevent many of the adverse outcomes identified in this work among initiators and persistent users, reducing the significant morbidity, mortality and cost associated with NUPM. Role of the funding sources The NESARC is funded by the National Institute on Alcoholism and Alcohol Abuse. NIAAA had no further role in study design, the collection, analysis or interpretation of data, the writing of the report, or the decision to submit the paper for publication. Jahn Hakes is employed by the US Census Bureau. Any views expressed
on statistical and technical issues are those of the author(s) and not necessarily those of the U.S. Census Bureau. Contributors Ty Schepis was the primary writer of the manuscript. Jahn Hakes conducted all statistical analyses. Both authors participated equally in devising the research plan, selecting statistical analyses and in editing the manuscript. Both authors have read and approve the final version of the manuscript. Conflicts of interest The authors note no conflicts of interest. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.drugalcdep.2014.06.009. References Adler, N.E., Rehkopf, D.H., 2008. U.S. disparities in health: descriptions, causes, and mechanisms. Annu. Rev. Public Health 29, 235–252. American Psychiatric Association, 2000. Diagnostic And Statistical Manual Of Mental Disorders: DSM-IV-TR. American Psychiatric Association, Washington, DC. Best, D., Savic, M., Beckwith, M., Honor, S., Karpusheff, J., Lubman, D.I., 2013. The role of abstinence and activity in the quality of life of drug users engaged in treatment. J. Subst. Abuse Treat. 45, 273–279. Boyd, C.J., McCabe, S.E., Cranford, J.A., Young, A., 2006a. Adolescents’ motivations to abuse prescription medications. Pediatrics 118, 2472–2480. Boyd, C.J., McCabe, S.E., Teter, C.J., 2006b. Medical and nonmedical use of prescription pain medication by youth in a Detroit-area public school district. Drug Alcohol Depend. 81, 37–45. Caldeira, K.M., O’Grady, K.E., Vincent, K.B., Arria, A.M., 2012. Marijuana use trajectories during the post-college transition: health outcomes in young adulthood. Drug Alcohol Depend. 125, 267–275. Centers for Disease Control and Prevention, 2012. CDC Grand Rounds: Prescription Drug OVERDOSES—A U.S. Epidemic. MMWR 61, 10–13. Gandek, B., Ware, J.E., Aaronson, N.K., Apolone, G., Bjorner, J.B., Brazier, J.E., Bullinger, M., Kaasa, S., Leplege, A., Prieto, L., Sullivan, M., 1998. Cross-validation of item selection and scoring for the SF-12 Health Survey in nine countries: results from the IQOLA Project. International Quality of Life Assessment. J. Clin. Epidemiol. 51, 1171–1178. Garnier, L.M., Arria, A.M., Caldeira, K.M., Vincent, K.B., O’Grady, K.E., Wish, E.D., 2009. Nonmedical prescription analgesic use and concurrent alcohol consumption among college students. Am. J. Drug Alcohol Abuse 35, 334–338. Gonzalez-Saiz, F., Rojas, O.L., Castillo, I.I., 2009. Measuring the impact of psychoactive substance on health-related quality of life: an update. Curr. Drug Abuse Rev. 2, 5–10. Goodwin, R.D., Hasin, D.S., 2002. Sedative use and misuse in the United States. Addiction 97, 555–562. Grant, B.F., Dawson, D.A., Stinson, F.S., Chou, P.S., Kay, W., Pickering, R., 2003a. The Alcohol Use Disorder and Associated Disabilities Interview Schedule-IV (AUDADIS-IV): reliability of alcohol consumption, tobacco use, family history of depression and psychiatric diagnostic modules in a general population sample. Drug Alcohol Depend. 71, 7–16. Grant, B.F., Harford, T.C., Dawson, D.A., Chou, P.S., Pickering, R.P., 1995. The Alcohol Use Disorder And Associated Disabilities Interview Schedule (AUDADIS): reliability of alcohol and drug modules in a general population sample. Drug Alcohol Depend. 39, 37–44. Grant, B.F., Kaplan, K., Shepard, J., Moore, T., 2003b. Source And Accuracy Statement For Wave 1 Of The 2001-2002 National Epidemiologic Survey On Alcohol and Related Conditions. National Institute on Alcohol Abuse and Alcoholism, Bethesda, MD. Grant, B.F., Kaplan, K.D., 2005. Source and Accuracy Statement for the Wave 2 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). National Institute on Alcohol Abuse and Alcoholism, Rockville, MD. Gugelmann, H., Perrone, J., Nelson, L., 2012. Windmills and pill mills: can PDMPs tilt the prescription drug epidemic? J. Med. Toxicol. 8, 378–386. Havens, J.R., Young, A.M., Havens, C.E., 2011. Nonmedical prescription drug use in a nationally representative sample of adolescents: evidence of greater use among rural adolescents. Arch. Pediatr. Adolesc. Med. 165, 250–255. Hertz, J.A., Knight, J.R., 2006. Prescription drug misuse: a growing national problem. Adolesc. Med. Clin. 17, 751–769. Katz, C., El-Gabalawy, R., Keyes, K.M., Martins, S.S., Sareen, J., 2013. Risk factors for incident nonmedical prescription opioid use and abuse and dependence: results
T.S. Schepis, J.K. Hakes / Drug and Alcohol Dependence 142 (2014) 161–167 from a longitudinal nationally representative sample. Drug Alcohol Depend. 132, 107–113. Kim, H.M., Smith, E.G., Ganoczy, D., Walters, H., Stano, C.M., Ilgen, M.A., Bohnert, A.S., Valenstein, M., 2012. Predictors of suicide in patient charts among patients with depression in the Veterans Health Administration health system: importance of prescription drug and alcohol abuse. J. Clin. Psychiatry 73, e1269–e1275. Kristman, V., Manno, M., Cote, P., 2004. Loss to follow-up in cohort studies: how much is too much? Eur. J. Epidemiol. 19, 751–760. Kuramoto, S.J., Chilcoat, H.D., Ko, J., Martins, S.S., 2012. Suicidal ideation and suicide attempt across stages of nonmedical prescription opioid use and presence of prescription opioid disorders among U.S. adults. J. Stud. Alcohol Drugs 73, 178–184. Manchikanti, L., Singh, A., 2008. Therapeutic opioids: a ten-year perspective on the complexities and complications of the escalating use, abuse, and nonmedical use of opioids. Pain Phys. 11, S63–S88. Martins, S.S., Fenton, M.C., Keyes, K.M., Blanco, C., Zhu, H., Storr, C.L., 2012. Mood and anxiety disorders and their association with non-medical prescription opioid use and prescription opioid-use disorder: longitudinal evidence from the National Epidemiologic Study on Alcohol and Related Conditions. Psychol. Med. 42, 1261–1272. Martins, S.S., Keyes, K.M., Storr, C.L., Zhu, H., Chilcoat, H.D., 2009. Pathways between nonmedical opioid use/dependence and psychiatric disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Drug Alcohol Depend. 103, 16–24. Martins, S.S., Keyes, K.M., Storr, C.L., Zhu, H., Grucza, R.A., 2010. Birth-cohort trends in lifetime and past-year prescription opioid-use disorder resulting from nonmedical use: results from two national surveys. J. Stud. Alcohol Drugs 71, 480–487. McCabe, S.E., Knight, J.R., Teter, C.J., Wechsler, H., 2005. Non-medical use of prescription stimulants among US college students: prevalence and correlates from a national survey. Addiction 100, 96–106. McCabe, S.E., West, B.T., Wechsler, H., 2007. Trends and college-level characteristics associated with the non-medical use of prescription drugs among US college students from 1993 to 2001. Addiction 102, 455–465. McCauley, J.L., Back, S.E., Brady, K.T., 2013. Pilot of a brief, web-based educational intervention targeting safe storage and disposal of prescription opioids. Addict. Behav. 38, 2230–2235. Oliva, E.M., Keyes, M., Iacono, W.G., McGue, M., 2012. Adolescent substance use groups: antecedent and concurrent personality differences in a longitudinal study. J. Pers. 80, 769–793. Rao, J.N.K., Scott, A.J., 1984. On chi-squared tests for multi-way tables with cell proportions estimated from survey data. Ann. Stat. 12, 46–60. Schepis, T.S., Hakes, J.K., 2011. Nonmedical prescription use increases the risk for the onset and recurrence of psychopathology: results from the National
167
Epidemiological Survey on Alcohol and Related Conditions. Addiction 106, 2146–2155. Schepis, T.S., Hakes, J.K., 2013. Dose-related effects for the precipitation of psychopathology by opioid or tranquilizer/sedative nonmedical prescription use: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J. Addict. Med. 7, 39–44. Schepis, T.S., Krishnan-Sarin, S., 2008. Characterizing adolescent prescription misusers: a population-based study. J. Am. Acad. Child Adolesc. Psychiatry 47, 745–754. Schulenberg, J., O’Malley, P.M., Bachman, J.G., Wadsworth, K.N., Johnston, L.D., 1996. Getting drunk and growing up: trajectories of frequent binge drinking during the transition to young adulthood. J. Stud. Alcohol 57, 289–304. Shedler, J., Block, J., 1990. Adolescent drug use and psychological health. A longitudinal inquiry. Am. Psychol. 45, 612–630. Simoni-Wastila, L., Strickler, G., 2004. Risk factors associated with problem use of prescription drugs. Am. J. Public Health 94, 266–268. Spoth, R., Trudeau, L., Shin, C., Ralston, E., Redmond, C., Greenberg, M., Feinberg, M., 2013. Longitudinal effects of universal preventive intervention on prescription drug misuse: three randomized controlled trials with late adolescents and young adults. Am. J. Public Health 103, 665–672. Spoth, R., Trudeau, L., Shin, C., Redmond, C., 2008. Long-term effects of universal preventive interventions on prescription drug misuse. Addiction 103, 1160–1168. Substance Abuse and Mental Health Services Administration, 2010. Drug Abuse Warning Network, 2007: Area Profiles of Drug-Related Mortality. Office of Applied Studies, HHS Publication No. SMA 09-4407, DAWN Series D-31, Rockville, MD. Substance Abuse and Mental Health Services Administration, 2012. Results from the 2011 National Survey on Drug Use and Health: Summary of National Findings. NSDUH Series H-44, HHS Publication No. (SMA) 12-4713. Substance Abuse and Mental Health Services Administration, Rockville, MD. Substance Abuse and Mental Health Services Administration; Office of Applied Studies, 2009. Treatment Episode Data Set (TEDS) Highlights - 2007 National Admissions to Substance Abuse Treatment Services. OAS Series #S-45, HHS Publication No. (SMA) 09-4360, Rockville, MD. Tucker, J.S., Ellickson, P.L., Collins, R.L., Klein, D.J., 2006. Are drug experimenters better adjusted than abstainers and users? A longitudinal study of adolescent marijuana use. J. Adolesc. Health 39, 488–494. Ugochukwu, C., Bagot, K.S., Delaloye, S., Pi, S., Vien, L., Garvey, T., Bolotaulo, N.I., Kumar, N., Ishak, W.W., 2013. The importance of quality of life in patients with alcohol abuse and dependence. Harv. Rev. Psychiatry 21, 1–17. Ware Jr., J., Kosinski, M., Keller, S.D., 1996. A 12-Item short-form health survey: construction of scales and preliminary tests of reliability and validity. Med. Care 34, 220–233.
