Substance Use and Addiction Research: Methodology, Mechanisms, and Therapeutics

By Alan David Kaye and Richard D. Urman

Substance Use and Addiction Research: Methodology, Mechanisms, and Therapeutics is an up-to-date, comprehensive, practical book on research methodologies for substance use and addiction that is intended for researchers and consumers of research information at all levels. The book is divided into four major sections, including an Introduction, Research Methodology for clinical trials, animal research and retrospective studies, Mechanisms of Use and Addiction, and Investigative Therapeutics: Designing and Measuring Outcomes. It serves a source for addressing all aspects of research design, methods and analysis within the context of the field of opioids, alcohol and other substances.

The book covers what is known in the field of quantitative and qualitative research methods, provides future directions, and introduces new models for investigation. It is organized around a translational science framework, with the contents addressing substance use/addiction research in the context of epidemiology, etiology, intervention efficacy and effectiveness, and implementation of evidence-informed interventions.

• Presents a practical, easy to read text designed to appeal to both experienced and beginner researchers in the field of substance abuse/addiction science
• Provides a concise, well-organized handbook that is a complete guide to methodologies in conducting substance abuse/addiction research
• Contains contributions from leading academic institutions
• Includes ample diagrams, tables and figures to help organize the information for easy reference, along with a list and explanation of existing useful measurement tools, websites, statistical methods and other resources

• Language: English
• Publisher: Academic Press
• Release date: Feb 10, 2023
• ISBN: 9780323986274

Book preview

Part 1

Research methodology for clinical trials, animal research, and retrospective studies

Outline

Chapter 1 Reliability and validity in substance misuse and addiction research

Chapter 2 Animal models

Chapter 3 Translational research strategies

Chapter 4 Experimental designs for addiction research

Chapter 5 Experimental design in clinical trials

Chapter 6 Qualitative and quantitative research methods

Chapter 7 Ethical issues in substance misuse and addiction-related research

Chapter 8 Informatics

Chapter 9 Artificial intelligence and machine learning

Chapter 10 Methods for data collection and analysis in epidemiology-of-substance-use research

Chapter 11 Utility of neuroimaging for substance misuse and addiction research: methodology, mechanisms, and therapeutics

Chapter 12 Systematic review and meta-analysis

Chapter 1

Reliability and validity in substance misuse and addiction research

Allyson L. Spence¹, Ndeloh Fontem² and Christine Feltman²,    ¹Department of Pharmaceutical, Social, and Administrative Sciences, Belmont University College of Pharmacy, Nashville, TN, United States,    ²Department of Pharmacy Practice, Regis University School of Pharmacy, Denver, CO, United States

Abstract

As the prevalence of substance use disorders continues to increase in the United States and globally, researchers and medical providers must identify effective tools to diagnose and treat addiction. The Addiction Severity Index (ASI) is commonly used to diagnose the severity of an individual’s addiction. Animal models of addiction also provide an important preclinical tool for studying addiction. Various reliability and validity measures provide insight into the effectiveness of these tools and paradigms. Reliability describes how well these tools can produce consistent results in varying circumstances. Common types of reliability include test–retest reliability, interrater reliability, and internal consistency. Both the ASI and many animal models of addiction provide high reliability as they provide consistent results when assessing addiction. Validity describes how well a specific procedure, tool, or questionnaire will measure what it intends to measure. Face, content, construct, criterion, and predictive validity are types of validity that can assess if these tools and paradigms measure what they intend to measure. Similar to reliability, both the ASI and animal models of addiction have demonstrated high validity.

Keywords

Reliability; validity; substance use disorder; Addiction Severity Index; addiction research; substance misuse research

Introduction

Substance use disorder has been on the rise in recent years. It presents both the individual and society with multiple harmful consequences, including impacts on the individual’s health and well-being and a financial burden on society. 2020 was a hallmark year in substance use disorders as many states experienced an all-time high in the number of drug overdoses treated within their emergency departments [1]. Animal models are an important preclinical tool for researchers as they try to identify more effective treatment options for individuals suffering from substance use disorders. These paradigms must present high reliability and validity for their results to translate well into humans [2]. This chapter outlines the reliability and validity of these animal models in studying addiction.

It is equally important that medical providers identify reliable and valid diagnostic tools to appropriately diagnose substance use disorders and their varying severities to implement appropriate treatment strategies [3,4]. The Addiction Severity Index (ASI) is commonly used in clinical settings to assess and diagnose an individual’s addiction to one or more substances and identify appropriate treatment strategies. This 1-hour interview is administered by a trained clinician who asks questions focused on multiple self-reported areas related to addiction: substance use, physical health, criminal activity, employment and financial support, relationships, and psychiatric disorders and/or symptoms. Over the past 30 years, research has demonstrated high reliability and validity of this assessment tool, thus strengthening its use [5,6]. This chapter emphasizes the reliability and validity of the ASI and its potential use in medical practice.

Reliability

Reliability describes how well a specific procedure, tool, or questionnaire will produce consistent results in various circumstances. If this procedure can consistently produce similar results under different circumstances (e.g., different persons administering the same questionnaire), it is considered reliable [7]. There are multiple measures of reliability, including test–retest reliability, interrater reliability, and internal consistency.

Test–retest reliability

To measure test–retest reliability, the goal is to conduct the same test for the same subjects in the group at two different points in time. Then you calculate the correlation of the sets of data. Having two different points in time will help assure there is low variability in the quality of data coming from the subjects [8]. Behavioral training and testing in animal models for addiction provide a high test–retest reliability. For example, self-administration requires consistent, reliable responses (i.e., less than 10% variation for three consecutive days) from subjects before testing periods can begin. Tests are repeated multiple times to eliminate outliers and ensure the test can reliably produce consistent results (i.e., if responses for drug self-administration are reduced 50% for one test, they should be similarly reduced for a repeat test) [9].

In terms of testing the reliability of the ASI via the test–retest method, many studies using the ASI have determined that short-term test–retest results show inconsistent results. In contrast, some studies suggest excellent reliability and others show unsatisfactory reliability. Some shortcomings of this reliability test were prominent in a few different testing areas, including severity ratings, days with medical problems, family or self-income, depression status, and composite scores (CSs) of medical, legal, and drug use. These limitations were some of the reasons for the instability in the test–retest reliability [5].

Interrater reliability

Interrater reliability is how two or more individual researchers conclude the same results when looking at the same testing population. This reliability is measured using a correlation coefficient, meaning if there is a high correlation coefficient, the researchers are very likely to produce the same results [10]. Interrater reliability is not typically measured for animal models of addiction, as the same researcher will ideally perform the research throughout the experiment. This is because most animals require an acclimation period where they acclimate to their environment and their handler before experimentation. For this reason, changing the primary scientist will likely require a new acclimation period before interrater reliability can accurately be measured [11].

In a study to determine the reliability and validity of the ASI, the researchers measured interviewer severity ratings (ISRs) and CSs. The ISRs first determined the severity of the need for additional treatment, and an estimate was made based on the subject’s responses on a 10-point scale. These data were then adjusted to the subject’s reflection on current addiction problems and the need for further treatment. With all these data points, the interrater reliability showed high reliability in using the ASI for severity ratings. However, there was a low-reliability coefficient in the problem areas (e.g., the effect of addiction on employment, drugs, family, and social areas deeming the results to be unstable) [5].

Internal consistency

Internal consistency gauges how reliably a test or survey measures the desired outcome. For example, if a survey includes two opposing true/false questions, such as I continue to use drugs in excess, despite its detrimental consequences on my personal and professional relationships and I am consistently capable of ceasing the use of drugs when I notice it damaging my personal and professional relationships. To demonstrate internal consistency, the individual would answer true to one of these questions and false to the other. Internal consistency cannot be easily measured in animal models of addiction, as they cannot be directly questioned on what they are feeling or experiencing [9].

Measuring internal consistency is far easier to measure in humans, and thus is very useful for the ASI since the goal is to determine how well this index measures both the ISR and CS. Within four different studies of internal consistency of the seven domains of CSs (i.e., medical status, alcohol use, psychiatric status, employment status, drug use, legal status, and relationships), strong internal consistency was found for only three domains (i.e., medical status, alcohol use, and psychiatric status) while low internal consistency was found for the remaining four domains [6].

Validity

Validity describes how well a specific procedure, tool, or questionnaire will measure what it intends to measure. For example, if a questionnaire intends to measure if an individual has a substance use disorder, it is not helpful if it measures anxiety, which would deem the questionnaire invalid [12].

Multiple types of validity are commonly used to evaluate procedures, including face validity, content validity, construct validity, criterion validity, and predictive validity. When using animal behavioral models in a preclinical setting, it is important to consider their validity. Three types of validity are commonly considered for animal models, including face validity, construct validity, and predictive validity. Face validity indicates that the behaviors of the model closely resemble the behaviors of the disease (i.e., substance use disorder), construct validity refers to the similarities in mechanisms underlying these behaviors. Predictive validity reflects the model’s ability to identify drugs with potential therapeutic value in humans [13].

Face validity

Face validity is a way to measure how a procedure appears effective to determine what it is set to measure. Many animal models of addiction present a high face validity, including the drug self-administration model, which is used to examine the reinforcing properties of a drug. In this model, animals are required to elicit a response (e.g., lever press) to obtain the drug (e.g., via an infusion) [14]. Reinforcement is defined as the ability of a drug to increase the probability of a response that precedes its delivery [15]. Thus a reinforcing drug would significantly increase the number of lever presses when compared to vehicle. The self-administration model offers a high degree of face validity, and animals will self-administer almost every drug misused by humans, with a few exceptions. For example, both rats and nonhuman primates self-administer opioids, cocaine, and methamphetamine [16].

However, some drugs of misuse have proven more difficult to demonstrate self-administration in rats and nonhuman primates, such as the cannabinoid Δ9-tetrahydrocannabinol and hallucinogens [17]. Therefore the face validity of the self-administration model should be addressed individually for drugs of misuse before using this model in a preclinical setting. Genetic factors, such as strain and sex, should also be considered when assessing the validity of the self-administration models since some rat strains, such as the Fisher 344 rats, will not self-administer drugs of misuse [18].

The reinstatement model, which involves periods of self-administration and/or drug exposure, extinction (i.e., removal of the drug), and drug-seeking behavior, is extensively used to study the relapse to drug addiction. Before relapse occurs, a drug-taking subject stops taking drugs, which is represented by extinction in the reinstatement model. Relapse occurs when the individual returns to the previously reinforced behavior with a drug injection, which is represented by lever pressing upon the presentation of drug-associated cues in the reinstatement model. Thus the animal model of reinstatement appears to have great face validity [19].

For the ASI to demonstrate a high face validity, it needs to accurately measure the severity of addiction an individual has. The ASI questionnaire has high face validity, which likely contributes to its frequent assessment of addiction [5,20].

Content validity

Content validity is a way to measure if a given test’s content accurately covers all the topics of a given interest. To exhibit a high content validity, the questions within the ASI should cover all the aspects of addiction. They should be able to measure the extent of addiction by the answers provided by the subject. The content of the ASI aligns with the criteria outlined in The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), thus giving this questionnaire a high content validity [6].

Construct validity

Construct validity measures the extent to which a given hypothesis is consistent with the intended use. It seeks to uncover if the test or tool measures its intended theoretical construct. To test for the construct validity of addiction, it is necessary to assess whether the subject has an addiction. We can measure a person’s cravings, risky behaviors, and withdrawal from drug use to validate that the individual likely suffers from a substance use disorder. Construct validity ensures that testing used for cravings and withdrawal is correlated with addiction [21].

The construct validity of animal models of addiction has only recently grown in interest. While the pathophysiological processes underlying substance use disorders are largely unknown, studies have suggested that repeated misuse of addictive substances leads to dysregulation of the brain rewards system, such as the mesocorticolimbic dopaminergic system [22]. Brain imaging studies in humans have shown that as dopamine levels increase in these brain regions (e.g., nucleus accumbens), the self-reported feeling of euphoria also increases. Similarly, increased withdrawal symptoms are associated with significant reductions in dopamine release and dopamine receptors [23]. Animal models of addiction result in similar alterations, such as increased dopamine during periods of reinforcement and decreased dopamine transporters and receptors upon chronic drug use, thus strengthening their construct validity [24].

Criterion validity

The criterion validity for ASI measures how much the test subject’s scores correlate with the criterion in the real world (or prediction of the outcome of another measure). For example, if a subject shows to have a high severity of addiction, their severity should directly correlate with their ability to function daily. However, criterion validity is expressed in different degrees of association. How close the association must be to be determined significant is based on the different variables and the subject of the study. The ASI demonstrates a high criterion validity and has shown high sensitivity and specificity in identifying the severity of substance use disorders when comparing the subject scores to the criteria outlined in the DSM-5 for the diagnosis of a substance use disorder [5].

Predictive validity

Animal models are frequently assessed for predictive validity, which refers to whether the observable effects of the disease in animals represent the disease in humans, particularly as it relates to treatment interventions. If a treatment effectively reduces the symptoms of substance use disorders in an animal model, it must be similarly efficacious in treating addictions in humans to have high predictive validity. The number of pharmacological treatment options for individuals suffering from addiction is limited, but of these limited options, many have strengthened the predictive validity of animal models of addiction. Naltrexone is an opioid receptor antagonist that reduced alcohol preference in animal models of addiction (i.e., hamsters). Thus it continued into clinical trials where it similarly reduced the pleasurable feelings associated with alcohol use, thus decreasing its craving and use in humans [25,26]. These studies led to the U.S. Food and Drug Administration approval of naltrexone to treat alcohol use disorders [27].

The ASI psychiatric CS has significant predictive validity and is useful in identifying patients with drug use disorders who could benefit from additional mental health treatment. This is especially important as substance use disorders have a high relapse rate, and identifying patients with increased severity of addiction is important for providing immediate and effective treatment options [28].

References

1. Zhang X, Wang N, Hou F, et al. Emergency department visits by patients with substance use disorder in the United States. West J Emerg Med. 2021;22(5):1076–1085.

2. Fuchs RA, Higginbotham JA, Hansen EJ. Animal models of addiction. Neural mechanisms of addiction Elsevier 2019;3–22 https://linkinghub.elsevier.com/retrieve/pii/B9780128122020000026.

3. Díaz Batanero C, Sanchez-Garcia M, Fernández-Calderón F, Lozano OM. Decision making in substance use disorder patients: reliability, convergent validity and sensitivity to change of a new behavioral measure. J Clin Exp Neuropsychol. 2020;42(10):998–1009.

4. Nelson CB, Rehm J, Ustün TB, Grant B, Chatterji S. Factor structures for DSM-IV substance disorder criteria endorsed by alcohol, cannabis, cocaine and opiate users: results from the WHO reliability and validity study. Addict Abingdon Engl. 1999;94(6):843–855.

5. Mäkelä K. Studies of the reliability and validity of the Addiction Severity Index. Addict Abingdon Engl. 2004;99(4):398–410 ; discussion 411–418.

6. Ljungvall H, Persson A, Åsenlöf P, Heilig M, Ekselius L. Reliability of the Addiction Severity Index self-report form (ASI-SR): a self-administered questionnaire based on the Addiction Severity Index composite score domains. Nord J Psychiatry. 2020;74(1):9–15.

7. Ahmed I, Ishtiaq S. Reliability and validity: importance in medical research. JPMA J Pak Med Assoc. 2021;71(10):2401–2406.

8. Collins LM. Research design and methods. Encyclopedia of gerontology Elsevier 2007;433–442 https://linkinghub.elsevier.com/retrieve/pii/B0123708702001621.

9. Spanagel R. Animal models of addiction. Dialogues Clin Neurosci. 2017;19(3):13.

10. Armstrong D, Gosling A, Weinman J, Marteau T. The place of inter-rater reliability in qualitative research: an empirical study. Sociology. 1997;31(3):597–606.

11. Rowan AN. Refinement of animal research technique and validity of research data. Fundam Appl Toxicol J Soc Toxicol. 1990;15(1):25–32.

12. Mohamad MM, Sulaiman NL, Sern LC, Salleh KM. Measuring the validity and reliability of research instruments. Procedia Soc Behav Sci. 2015;204:164–171.

13. Pechnick RN, Glasner-Edwards S, Hrymoc M, Wilkins JN. Preclinical development and clinical implementation of treatments for substance abuse disorders. FOCUS. 2007;5(2):151–162.

14. Clark R, Schuster CR, Brady JV. Instrumental conditioning of jugular self-infusion in the rhesus monkey. Science. 1961;133(3467):1829–1830.

15. Balster RL, Bigelow GE. Guidelines and methodological reviews concerning drug abuse liability assessment. Drug Alcohol Depend. 2003;70(3):S13–S40.

16. O’Connor EC, Chapman K, Butler P, Mead AN. The predictive validity of the rat self-administration model for abuse liability. Neurosci Biobehav Rev. 2011;35(3):912–938.

17. Panlilio LV, Goldberg SR. Self-administration of drugs in animals and humans as a model and an investigative tool. Addict Abingdon Engl. 2007;102(12):1863–1870.

18. Suzuki T, George FR, Meisch RA. Etonitazene delivered orally serves as a reinforcer for Lewis but not Fischer 344 rats. Pharmacol Biochem Behav. 1992;42(4):579–586.

19. Katz JL, Higgins ST. The validity of the reinstatement model of craving and relapse to drug use. Psychopharmacology (Berl). 2003;168(1–2):21–30.

20. Luo W, Wu Z, Wei X. Reliability and validity of the Chinese version of the addiction severity index. J Acquir Immune Defic Syndr. 2010;53(Suppl 1):S121–S125.

21. Field M, Kersbergen I. Are animal models of addiction useful?. Addict Abingdon Engl. 2020;115(1):6–12.

22. Baik J-H. Dopamine signaling in reward-related behaviors. Front Neural Circuits 2013;:7 http://journal.frontiersin.org/article/10.3389/fncir.2013.00152/abstract.

23. Volkow ND, Fowler JS, Wang G-J, Swanson JM, Telang F. Dopamine in drug abuse and addiction: results of imaging studies and treatment implications. Arch Neurol. 2007;64(11):1575.

24. Willuhn I, Wanat MJ, Clark JJ, Phillips PEM. Dopamine signaling in the nucleus accumbens of animals self-administering drugs of abuse. Curr Top Behav Neurosci. 2010;3:29–71.

25. Ross D, Hartmann RJ, Geller I. Ethanol preference in the hamster: effects of morphine sulfate and naltrexone, a long-acting morphine antagonist. Proc West Pharmacol Soc. 1976;19:326–330.

26. Volpicelli JR. Naltrexone in the treatment of alcohol dependence. Arch Gen Psychiatry. 1992;49(11):876.

27. Sudakin D. Naltrexone: not just for opioids anymore. J Med Toxicol J Am Coll Med Toxicol. 2016;12(1):71–75.

28. Thylstrup B, Bloomfield K, Hesse M. Incremental predictive validity of the Addiction Severity Index psychiatric composite score in a consecutive cohort of patients in residential treatment for drug use disorders. Addict Behav. 2018;76:201–207.

Chapter 2

Animal models

Allyson L. Spence¹, Courtney M. Keller², Maggie Mott² and Kevin S. Murnane², ³, ⁴,    ¹Department of Pharmaceutical, Social, and Administrative Sciences, Belmont University College of Pharmacy, Nashville, TN, United States,    ²Department of Pharmacology, Toxicology & Neuroscience, Louisiana State University Health Sciences Center at Shreveport, Shreveport, LA, United States,    ³Department of Psychiatry and Behavioral Medicine, School of Medicine, Louisiana State University Health Sciences Center at Shreveport, Shreveport, LA, United States,    ⁴Louisiana Addiction Research Center, Louisiana State University Health Sciences Center at Shreveport, Shreveport, LA, United States

Abstract

Animal models provide an important tool for studying addiction/substance use disorders. These animal models mimic the criteria outlined by the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition for substance use disorders, strengthening support for their use to help researchers better understand the underlying mechanisms of substance use disorders and potential treatment strategies for treating these disorders. Commonly used animal models of addiction include self-administration, conditioned place preference, and drug discrimination. Self-administration, which measures the reinforcing properties of a drug, closely mimics how humans use these substances, as the animal can control how much drug they consume. Conditioned place preference measures the rewarding properties of a drug and allows researchers to measure extinction and reinstatement of the conditioned substance. Drug discrimination, which measures the subjective effects of a drug, requires animals to discriminate between drug and nondrug conditions.

Keywords

Substance use disorder; animal models of addiction; self-administration; reinforcement; conditioned place preference; drug discrimination

Introduction

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (i.e., DSM-V), which is considered the gold standard for diagnosing mental health disorders, defines substance use disorders as a chronic, compulsive condition in which the individual continues to take a substance despite its negative effects [1]. Decades of research have allowed the DSM-V to outline the 11 criteria for substance use disorder, emphasizing physical dependence, risky use, impaired control, and social problems that arise from substance use disorders. Commonly misused substances within the United States include alcohol, tobacco, stimulants, opioids, and hallucinogens. Animal models provide a valuable tool for studying these substance use disorders so that researchers can better understand the pathophysiological characteristics underlying this disorder and identify potential treatments [2]. These animal models have a high degree of predictive, face, and construct validities, all of which strengthen the support for their use [3].

The addiction cycle is credited with having three primary stages: the binge/intoxication stage, where the individual consumes the substance to enjoy its euphoric effects; the withdrawal/negative affect stage, where the individual experiences unpleasant withdrawal symptoms upon cessation of drug use; and the preoccupation/anticipation stage, where the individual craves the drug and may encounter triggers that remind of them of their prior substance use and often lead to relapse to drug use [4]. Various animal models of addiction allow us to study these different stages of addiction. For instance, self-administration allows us to study the binge/intoxication stage as we observe the animal increase and eventually stabilize their self-administration of a substance [5]. This chapter outlines some of the most common animal models of addiction, including self-administration, conditioned place preference, and drug discrimination.

Self-administration

Self-administration procedures are used to study the reinforcing effects of drugs in a controlled environment. These studies are typically done using rodent or nonhuman primates but have also been conducted using dogs, pigeons, and even invertebrate animals, including honeybees, crayfish, and ants [6–12]. Self-administration procedures are based on the principles of operant conditioning. In these studies, animals are trained to perform a specific action (e.g., lever press or nose poke) to obtain a drug. Since drugs of misuse are powerful reinforcers, animals will continue to perform the specified behavior to receive the drug of misuse (i.e., the reinforcer). A drug is considered reinforcing if the behavior leading to its delivery increases over time [13]. Self-administration procedures have high predictive validity as animals will reliably self-administer almost all drugs misused by humans [11].

Intravenous drug self-administration

Intravenous drug self-administration is the standard procedure used to model human substance use disorder. This procedure was developed in the 1960s using nonhuman primates and rats [14,15]. In these studies, animals are infused intravenously with the drug of misuse contingent upon the subject emitting a specific behavior. Thus delivery of the drug to the central nervous system is rapid. An indwelling catheter must be implanted via the jugular vein or femoral vein before the experiment can begin. Thus catheter patency can limit the duration of the experiment. Another limitation of intravenous self-administration studies is the solubility of a compound, as less soluble compounds may not be suitable for intravenous delivery.

Oral self-administration

Oral self-administration has been used to study the reinforcing effects of multiple drugs, including nicotine and opioids, but is primarily used to study alcohol use disorder. Since humans typically misuse alcohol via the oral route of administration, this model has clear face validity. In operant oral self-administration studies, performing a specific response results in the availability of a drug via a cup or a sipper tube. The sipper tube method is advantageous over the cup since it allows for measuring consumed drug. Operant oral self-administration studies can be challenging due to the aversive taste of some compounds, including alcohol. To overcome this issue, sweetener (e.g., saccharin or sucrose) can be added to the drinking solution to encourage drug self-administration [16]. The concentration of sweetener added to the drug solution is gradually faded out throughout operant training. While adding sweetener to the drug solution can promote drug-taking behavior, it also adds a potential confound to the study. For example, sweeteners may affect the pharmacokinetics of alcohol [17].

In addition to operant oral self-administration studies, oral self-administration studies also include the two-bottle choice paradigm. In this procedure, animals are given two bottles of two different solutions (e.g., a drug solution and a water solution) and allowed to drink freely from both bottles [18]. The preference for the drug solution is then determined based on the total intake from both bottles. Unfortunately, the free choice between the two solutions may not result in a high enough consumption of the drug solution to produce intoxicating blood alcohol concentrations [19]. Because of this, the two-bottle choice method has been adapted to better model clinical aspects of substance use disorder. For example, drinking-in-the-dark and intermittent-access models have been developed to model binge drinking. In the drinking-in-the-dark model, rodents are given limited access (2–4 hours) to ethanol during the animal’s dark cycle [20]. This model is advantageous over the two-bottle choice procedure as it results in blood alcohol concentrations that are pharmacologically relevant [20]. High alcohol consumption can also be achieved by using the intermittent-access model, in which rodents are offered an alcohol solution every other day [21,22].

Intracranial self-administration

In the intracranial self-administration paradigm, animals are implanted stereotaxically with a cannula aimed at a specific brain region. Once the subject recovers from surgery, the animal can begin self-administration. The drug is infused directly into the implanted brain region when the animal produces the correct response. This procedure has been used to determine discrete brain regions that mediate the reinforcing effects of multiple drugs of misuse, including alcohol, cocaine, opioids, and amphetamine [23–26].

Operant vapor self-administration

Operant vapor self-administration has been developed to study drugs misused via the inhalation route of administration, such as nicotine and cannabis [27,28]. Since these drugs are commonly misused via inhalation, operant vapor self-administration has greater face validity than the intravenous self-administration paradigm for these drugs of misuse. In these experiments, subjects are trained to produce a specific response to receive a puff of drug vapor.

Schedules of reinforcement

Ratio and interval schedules of reinforcement

Schedules of reinforcement are used in self-administration studies to define the criteria that must be met for a behavior to be reinforced. In other words, a schedule of reinforcement is a rule that must be followed for the animal to receive the reinforcer (i.e., the drug of misuse). Most reinforcement schedules can be classified as either a ratio schedule or an interval schedule. In ratio schedules of reinforcement, an animal must perform a behavior a specific number of times to receive the reinforcer. In interval schedules of reinforcement, a behavior is only reinforced if it is performed after a set amount of time has passed. Both types of schedules can be either fixed schedules or variable schedules of reinforcement. In a fixed-ratio (FR) schedule of reinforcement, the number of responses that must occur before the reinforcer is delivered remains the same throughout the session. For example, in a FR 1 schedule, the reinforcer will be delivered after each response. In a variable-ratio schedule, the number of responses required to receive a reinforcer will vary throughout the session. In a fixed-interval (FI) schedule, the interval of time remains the same throughout the session. For example, in a FI 1-minute schedule, the first response after 1-minute has passed will be reinforced. The next reinforcer will be delivered only if the animal makes a response after one minute has elapsed. Responses made during the interval are not reinforced. Animals on a variable-interval (VI) schedule are reinforced for responses made after an unpredictable amount of time has passed.

Interval schedules of reinforcement and FR of reinforcement result in different self-administration behavioral patterns. In ratio schedules of reinforcement, the animal can increase the number of reinforcers received by performing the specified behavior more often; therefore the rate of reinforcement is directly related to the rate of responding. Because of this, ratio schedules result in a high, steady rate of responding throughout the session. In interval schedules of reinforcement, the number of reinforcer presentations is limited by the length of the intervals. For example, a subject undergoing a 1-hour session on a FI 10-minute schedule will receive a maximum of six presentations of the reinforcer, regardless of how high the rate of responding is. Thus the rate of responding and the reinforcer rate are not correlated in interval schedules. Because reinforcer delivery is limited, FI schedules of reinforcement produce a characteristic scalloped pattern of behavior where animals respond more toward the end of the interval and then slow their responding after the reinforcer is delivered [29]. VI schedules, on the other hand, result in a more constant rate of responding since the interval length is varied throughout the session [29].

Second-order schedules of reinforcement

In second-order schedules of reinforcement, subjects respond for the reinforcer itself and stimuli that have previously been paired with the reinforcer (i.e., conditioned reinforcers). For example, conditioned stimuli may be presented on a FR schedule while the drug is delivered based on a FI schedule of reinforcement. Second-order schedules maintain high rates of responding even when drug delivery rates are low. Second-order schedules of reinforcement offer a distinct advantage over ratio and interval schedules because second-order schedules can be used to measure drug-taking behavior and drug-seeking behavior [30].

Progressive-ratio schedules of reinforcement

A progressive-ratio schedule of reinforcement requires an animal to increase the number of responses for successive presentations of the reinforcer [31]. For instance, the first reinforcer may be delivered after one response, whereas the next presentation of the reinforcer may occur after five responses. This schedule of reinforcement is used to determine the breakpoint, which is defined as the largest ratio that the animal will complete for the reinforcer [31]. The breakpoint is often used to measure the animal’s motivation to take the drug and can be altered by several variables, including the dose of the drug and pharmacological manipulations. Increasing the dose of a self-administered drug increases the breakpoint, while decreasing the dose lowers the breakpoint [32,33]. However, higher doses of drugs have also been shown to decrease the breaking point, resulting in an inverted U-shaped dose–response curve [34,35]. Pharmacological compounds can also affect the breakpoint of a drug of misuse. For example, dopamine receptor antagonists lower the breakpoint of cocaine [32,36]. Overall, the progressive-ratio schedule of reinforcement is an effective tool for assessing the reinforcing efficacy of