Analyses related to the development of DSM-5 criteria for substance use related disorders: 1. Toward amphetamine, cocaine and prescription drug use disorder continua using Item Response Theory

https://doi.org/10.1016/j.drugalcdep.2011.09.004Get rights and content

Abstract

Background

Prior research has demonstrated the dimensionality of alcohol, nicotine and cannabis use disorders criteria. The purpose of this study was to examine the unidimensionality of DSM-IV cocaine, amphetamine and prescription drug abuse and dependence criteria and to determine the impact of elimination of the legal problems criterion on the information value of the aggregate criteria.

Methods

Factor analyses and Item Response Theory (IRT) analyses were used to explore the unidimensionality and psychometric properties of the illicit drug use criteria using a large representative sample of the U.S. population.

Results

All illicit drug abuse and dependence criteria formed unidimensional latent traits. For amphetamines, cocaine, sedatives, tranquilizers and opioids, IRT models fit better for models without legal problems criterion than models with legal problems criterion and there were no differences in the information value of the IRT models with and without the legal problems criterion, supporting the elimination of that criterion.

Conclusion

Consistent with findings for alcohol, nicotine and cannabis, amphetamine, cocaine, sedative, tranquilizer and opioid abuse and dependence criteria reflect underlying unitary dimensions of severity. The legal problems criterion associated with each of these substance use disorders can be eliminated with no loss in informational value and an advantage of parsimony. Taken together, these findings support the changes to substance use disorder diagnoses recommended by the American Psychiatric Association's DSM-5 Substance and Related Disorders Workgroup.

Introduction

One major change to substance use disorder diagnoses appearing in the Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition (DSM-5; American Psychiatric Association, 2011) will entail combining DSM-Fourth Edition (DSM-IV; American Psychiatric Association, 1994) abuse and dependence criteria to create one diagnosis for each substance assessed in the manual. Another change is the elimination of the legal problems of the DSM-IV abuse criteria as the result of its greater severity and much lower probability of endorsement relative to other abuse and dependence criteria.

Numerous studies using Item Response Theory (IRT) analyses have shown that DSM abuse and dependence criteria map well onto unidimensional continua of substance use disorder severity. IRT analysis defines the relationship between the observed responses to the criteria items and the underlying unobserved trait or construct (drug use disorder severity) and provides information of criterion functioning in terms of severity and discrimination. This result has been consistently robust across alcohol (Borges et al., 2010, Keys et al., 2010, Proudfoot et al., 2006, Saha et al., 2006, Saha et al., 2007, Shmulewitz et al., 2010), nicotine (McBride et al., 2010, Saha et al., 2010), cannabis (Breseler and Hasin, 2010, Compton et al., 2009, Martin et al., 2006, Teesson et al., 2002), amphetamine and cocaine (Gillespie et al., 2007, Langenbucher et al., 2004, Wu et al., 2009), hallucinogens and inhalant/solvent use (Kerridge et al., 2011) and illicit drug use disorders (Lynskey and Agrawal, 2007). Few studies, however, have examined the unidimensionality of amphetamine, cocaine and prescription drug use disorders in the general population. No study to our knowledge has explored the invariance of demographic characteristics on the latent construct nor directly examined the impact of eliminating the DSM-IV legal problems criterion associated with any substance except alcohol (Saha et al., 2006) and hallucinogens and inhalant/solvent use (Kerridge et al., 2011).

Among psychometric studies of cocaine use disorder criteria conducted in clinical settings (Langenbucher et al., 2004, Wu et al., 2009), sample sizes were small (<383) and in one of these studies (Wu et al., 2009) only dependence, but not abuse criteria were assessed. Another study focusing on the unidimensionality of both cocaine and amphetamines among Caucasian males participating in the Virginia Twin Registry, deliberately did not obtain information on abuse and dependence from at least 40% of the sample as the result of conditioning the assessment on restrictive definitions of substance use combined with requiring endorsement of at least one abuse criterion prior to assessment of dependence criteria (Gillespie et al., 2007). Yet, much less is known about the corresponding psychometric properties of sedative, tranquilizer and opioid use disorder criteria. Two studies, one focusing on adolescents (Wu et al., 2009) and the other on Caucasian Virginia Twins (Gillespie et al., 2007) found evidence to support the unidimensionality of sedatives and opioid abuse and dependence criteria, but the samples were limited and not generalizable to the full adult U.S. population. In another study conducted in the general population (Lynskey and Agrawal, 2007), discrimination and severity parameters of illicit drug abuse and dependence criteria were approximated from confirmatory factor analytic parameters and did not explore the impact of elimination of the legal problems criterion or invariance in factor scores.

In view of the methodological limitations of prior research, dearth of information on the impact of the elimination of the legal problems criteria from the substance use disorders, and strong evidence of the unidimensionality of abuse and dependence criteria for other substances including alcohol, cannabis and nicotine, as derived from Item Response Theory (IRT) analyses, the objectives of this study were to use IRT to: (1) determine whether amphetamine, cocaine, and prescription drugs (sedative, tranquilizer and opioid) abuse and dependence criteria measure unitary dimensions of severity in a large general population survey of the United States, the National Institute on Alcohol Abuse and Alcoholism's National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) (Grant et al., 2003); (2) determine the impact of removing the legal problems criterion on IRT model fit and informational value of the amphetamine, cocaine, sedatives, tranquilizer and opioid use disorder diagnoses; and (3) examine differential criterion functioning as a means to assess bias in the diagnoses across important subgroups of the population such as sex, age and race-ethnicity.

Section snippets

Sample

The 2001–2002 NESARC is a representative sample of the USA conducted by National Institute on Alcohol Abuse and Alcoholism (NIAAA) and described elsewhere (Grant et al., 2003, Grant et al., 2004). The NESARC target population was the civilian non-institutionalized population residing in households and selected group quarters, 18 years and older. Face-to-face interviews were conducted with 43,093 respondents, with a response rate of 81%. Blacks, Hispanics and young adults (ages 18–24) were

Prevalence and factor analyses

Lifetime prevalences of DSM-IV amphetamine abuse and dependence criteria ranged from 5.1% for the legal problems criterion to 38.2% for the quit/control criterion (Table 1). The one-factor solution was shown to be a good fit to the amphetamine criteria regardless of whether the model included (CFI = 0.989; TLI = 0.987; RMSEA = 0.042) or excluded (CFI = 0.989; TLI = 0.986; RMSEA = 0.047) the legal problems criterion.

Table 2 shows that lifetime prevalences of DSM-IV cocaine abuse and dependence criteria

Discussion

Similar to the results of the majority of research on alcohol (Borges et al., 2010, Keys et al., 2010, Saha et al., 2006, Saha et al., 2007), nicotine (McBride et al., 2010, Saha et al., 2010, Shmulewitz et al., 2010), cannabis (Compton et al., 2009, Teesson et al., 2002) and hallucinogens and inhalant/solvent use (Kerridge et al., 2011), both factor and IRT analyses demonstrated the utility of conceptualizing amphetamine, cocaine, sedatives, tranquilizer and opioid abuse and dependence

Role of funding source

The National Institute on Alcohol Abuse and Alcoholism (NIAAA) funded the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) with supplemental funding from the National Institute on Drug Abuse (NIDA).

Contributors

Dr. Saha conducted statistical analyses. Drs. Saha and Compton wrote the first draft of the manuscript. Drs. Grant, Chou and Smith revised and commented on subsequent drafts of the manuscript. Drs. Saha, Smith, Chou, Huang and Mr. Pickering and Ms. Ruan collected the data and administered the study.

Conflict of interest

There are no conflicts of interest.

Acknowledgments

One recommendation by the Editor of this Journal was to combine two separate manuscripts, one concerning amphetamines and cocaine and the other focusing on prescription drug use disorders. In view of this recommendation, this combined manuscript has two Co-First authors, Drs. Saha and Compton.

References (40)

  • D. Shmulewitz et al.

    The dimensionality of alcohol use disorders: results from Israel

    Drug Alcohol Depend.

    (2010)
  • M. Teesson et al.

    The structure of cannabis dependence in the community

    Drug Alcohol Depend.

    (2002)
  • L.T. Wu et al.

    The construct and measurement equivalence of cocaine and opioid dependences: a National Drug Abuse Treatment Clinical Trials Network (CTN) study

    Drug Alcohol Depend.

    (2009)
  • American Psychiatric Association

    Diagnostic and Statistical Manual of Mental Disorder, 5th ed. Draft Criteria

    (2011)
  • American Psychiatric Association

    Diagnostic and Statistical Manual of Mental Disorder

    (1994)
  • P.M. Bentler

    Comparative fit indexes in structural models

    Psychol. Bull.

    (1990)
  • R.D. Bock et al.

    Marginal maximum likelihood estimation of item parameters: application of an EM algorithm

    Psychometrika

    (1981)
  • K.A. Bollen

    Structural Equations with Latent Variables

    (1989)
  • D.M. Bolt et al.

    A multigroup item response theory analysis of the Psychopathy Checklist-revised

    Psychol. Assess.

    (2004)
  • G. Borges et al.

    The dimensionality of alcohol use disorders and alcohol consumption in a cross-national perspective

    Addiction

    (2010)
  • Cited by (0)

    View full text