The high prevalence of comorbid substance use disorders (SUDs) and mental illness (MI) has been well established in both clinical- and population-based studies (Brooner et al. 1997; Hien et al. 1997; Regier et al. 1990; Swendsen et al. 2010). The Epidemiological Catchment Area (ECA) study reported that among individuals with MI, 29 % had a comorbid SUD (compared with 13 % of individuals without a MI). Findings from the National Comorbidity Survey report a significantly higher prevalence of SUDs among individuals with MI than in the general population; among individuals with any MI, 51 % were reported as having a comorbid SUD, and the odds ratio for a SUD among individuals with a MI was 2.4 (compared to those without a MI) (Kessler et al. 1996a, b).

A recent study has explored the issue of transition from drug use to SUD using the National Epidemiological Survey of Alcohol and Related Conditions (NESARC) database. The NESARC database has been developed using face-to-face interviews on 43,093 adults representative of the civilian non-institutionalized population residing in the United States. The lifetime prevalence of any MI (mood, anxiety, psychotic, or personality disorder) in the NESARC sample was 33.7 %, while 66.2 % had no lifetime MI. Rates of transition from substance use to a SUD were calculated as the prevalence of lifetime SUD among individuals with lifetime exposure to a substance. The rate of transition from substance use to SUD was higher for individuals with a lifetime diagnosis of MI than for individuals without any MI. This was true for cocaine (OR = 1.83), cannabis (OR = 1.91), nicotine (OR = 3.24), alcohol (OR = 2.2), and hallucinogens (OR = 2.14) compared to controls (Lev-Ran et al. 2013). Interestingly, the SUD liability for nicotine was highest among the substances studied, in the range of 60 % for individuals with mood, anxiety, personality, and psychotic disorders (Lev-Ran et al. 2013).

Drug addiction also appears to be associated with high prevalence of tobacco use. Smoking rates ranging from 71 to 97 % were found among alcohol and other drug-dependent populations (Battjes 1988). Drug-abuse patterns seem to involve progression from one class of legal drug (alcohol or cigarettes) to marijuana/alcohol and eventually psychostimulants as well as opiate drugs (SAMHSA 2012; Yamaguchi and Kandel 1984). It has been noticed in the literature that two-thirds of drug abusers are regular tobacco smokers, a rate more than triple that of the rest of the population (Zickler 2000). In fact, early tobacco smoking has been proposed to constitute a “gateway” or rather a “common liability” (Vanyukov et al. 2012) toward substance abuse as has been suggested in the literature (Kandel and Faust 1975; Yamaguchi and Kandel 1984). Nicotine is an easily available drug for adolescents and might constitute in some cases the first contact with an addictive substance; however, some studies have pointed out that the reverse process or, as it has been denominated, the “reverse gateway theory,” might also happen. The reverse gateway theory suggests that earlier regular use of other substances, as cannabis, might predict later tobacco initiation and/or nicotine dependence in those who did not use tobacco before (Peters et al. 2012). An example illustrating this reverse gateway theory is the observation of substantial percentages of studied populations found to consume marijuana prior the use of licit drugs (Tarter et al. 2006). Both scenarios (i.e., gateway and reverse gateway theories) might result in a net increase on tobacco consumption, independently of tobacco being the initial abused substance or being initiated following other addiction . Due to the high rates of polydrug use of the alcohol/tobacco combination, it has been hypothesized in the literature the possibility of mutual potentiating effects of each of these legal substances on each other.

Evidence from multiple research studies supports the existence of developmental stages and sequences in drug use (Kandel 1975). The likelihood of first initiating tobacco or other legal drugs before using illegal drugs is much greater than the opposite process. There are reports showing progression from “soft” drugs to “hard” drugs on a 75–80 % of cases versus a 20–25 % progression from “hard” to “soft” drugs of the cases depending on the sample studied (George and Moselhy 2005; Tarter et al. 2006). Accordingly, in a recent study, it was found that first initiating tobacco appeared 17.6 times greater than the likelihood of initiating cannabis (Mayet et al. 2011). Similarly to the net increase in tobacco intake that might happen in the process of escalating drugs or the potentiating effects during polydrug use mentioned above, the reverse gateway theory proposes that net increases in tobacco or other legal drugs consumption might happen following the exposure to other illegal drugs. Recent epidemiological evidence shows that alcohol and illicit drug dependence are associated with increased risk for cigarette smoking (Redner et al. 2014). Other epidemiological studies have also shown cannabis use prior tobacco consumption in a number of cases (Agrawal et al. 2012; Tullis et al. 2003; Vaughn et al. 2008). The use of cannabis during young adulthood has also been associated with increased risk of later initiation of tobacco use and nicotine dependence (Patton et al. 2005). However, in other studies, the use of marijuana in adolescence has been found only modestly associated with daily cigarette smoking and nicotine dependence in young adulthood (Timberlake et al. 2007).

There are several possibilities as to why these comorbidities exist. Here, we will focus on the behavioral data generated using animal models that could explain the cause of those comorbidities. We will notably cover the evidence collected using animal models exploring how addictive disorders, schizophrenia , mood and anxiety disorders could facilitate nicotine addiction . It appears that the reinforcing/rewarding effects of nicotine are mediated through the nicotinic acetylcholine receptors (nAchRs). nAchRs are ligand-gated ion channels composed of five subunits, labeled α₂ to α₁₀ and β₂ to β₄ in the central nervous system. The combination of subunits determines the functionality of the receptor. In the brain, the most prevalent of these are the homo-oligomeric α₇ and the heteromeric α₄β₂^* nAChRs. The role and function of these combinations are not yet fully elucidated. There is clear evidence implicating α₄β₂^* nAChRs in nicotine addiction (Picciotto et al. 1998; Tapper et al. 2004). Midbrain α₄β₂^* nAChRs are critical for nicotine’s ability to increase dopamine levels in the nucleus accumbens (Corrigall 1991; Dani and De Biasi 2001; Laviolette and van der Kooy 2004; Picciotto and Corrigall 2002; Watkins et al. 2000), a feature that appears critical to mediate reinforcing/rewarding effects of nicotine (see chapter entitled The Role of Mesoaccumbens Dopamine in Nicotine Dependence; this volume). It appears that other nAChRs subunits such as α₅ (Fowler et al. 2011, 2013), α₆ (Pons et al. 2008), and α₇ (Besson et al. 2012) may be implicated as well.

Animal models of drug abuse may serve to test in a relatively simple manner both gateway and reverse gateway hypotheses. Indeed, those hypotheses predict that experience with one drug can enhance the addictive properties of a different drug. Thus, by exposing experimental animals to one drug and later evaluating changes in the addictive properties of a different drug (e.g., self-administration pattern) seems a straightforward manner to test these hypotheses.

The number of studies evaluating the effects of nicotine pre-exposure on the addictive properties of other drugs is clearly superior to the number of studies evaluating the opposite process. Only a few studies have explored the effect of pre-exposure to different drugs on nicotine’s-addictive properties. Thus, in a recent study, pre-exposure to tetrahydrocannabinol (THC) was able to increase the ratio of animals self-administering nicotine from 65 %, in vehicle-exposed rats, to 94 % in the THC-exposed animals (Panlilio et al. 2013). Moreover, rats pre-exposed to THC exhibited higher rates of nicotine taking over the course of acquisition training when compared to the control group. The substantial increase on the ratio of animals self-administering nicotine observed was likely by an increase in nicotine-reinforcing properties, as measured by a behavioral-economics procedure, where pre-exposure to THC produced a more persistent nicotine self-administration as the price (number of responses required) for nicotine raised (Panlilio et al. 2013). Previous studies have shown that acute pre-treatment with cannabinoid agonists and antagonists is able to modify other responses to nicotine as nicotine-induced changes in locomotor activity without changes in nicotine-evoked dopamine release (Kelsey and Calabro 2008; Rodvelt et al. 2007). Therefore, it would be interesting to further investigate the neurobiological correlates for nicotine response following pre-exposure to THC.

A recent study in mice found that pre-exposure to cocaine did not alter subsequent locomotor response to nicotine (Levine et al. 2011) (while nicotine pre-exposure facilitated subsequent responses to cocaine). In contrast, in male Sprague Dawley rats, animals receiving daily injections of cocaine for 14 days were more reactive to nicotine as compared to rats pretreated with a daily injection of saline. It should be noted that this effect was noted after 3 and 7 days of withdrawal from cocaine treatment (but not at the first day of withdrawal) (Szabo et al. 2014).

Preclinical studies have shown that a combination of legal drugs during early stages of development might also contribute to a later vulnerability to nicotine dependence. Thus, a combination of alcohol and nicotine during gestational periods might increase nicotine’s-reinforcing properties during adulthood (Matta and Elberger 2007). Interestingly, alcohol-naïve offspring of rats selectively bred for high alcohol intake exhibited increased vulnerability to nicotine self-administration and relapse (Le et al. 2006). Conversely, it was previously reported that rats voluntarily consume nicotine or alcohol independently of each other, and the pre-exposure to either nicotine or alcohol did not affect subsequent intake of both drugs in combination (Marshall et al. 2003). This issue has been explored more recently using procedure in which rats voluntarily self-administer nicotine , alcohol, or both at different time points of the experimental procedure (Le et al. 2010). It appears that the nicotine intake in overall was similar in animals that got previously access to alcohol and that were trained secondarily to self-administer nicotine , as compared to animals that were trained directly to self-administer nicotine without previous alcohol access (Le et al. 2010).

There are over 60 animal models of schizophrenia , and they all assess varying aspects of the disorder. Some of them model the positive symptoms, which include hallucinations and delusions. Others model the negative symptoms such as avolition and anhedonia, including changes in social functioning. Finally, some animal models study the cognitive symptoms such as impaired memory and attention. Very limited work has been conducted exploring the responses to nicotine on those models.

Depression-like phenotypes are most commonly assessed in rodents using the forced swim test or the sucrose preference test. In the forced swim test, rodents are placed in a beaker of water from which they cannot escape. At first, rodents will struggle and swim around. After a period, they will become immobile and float. Greater time spent immobile during the test is thought to reflect “despair” and the forced swim test is often used to screen for antidepressant medications as drugs with known antidepressant properties tend to decrease time spent immobile (Porsolt et al. 1978, 1977). The sucrose preference test is meant to measure anhedonia or decreased ability to find pleasure in normally pleasurable things. Rodents are given free choice between drinking from a bottle of water or a bottle of sucrose water available in the same cage. Usually, rodents show a strong preference for sucrose water; a significant decrease or lack of such a preference is interpreted as anhedonia (Katz 1982; Nestler and Hyman 2010).

Two behavioral tests have been used most frequently to model anxiety disorder-like states in rodents: the social interaction test and the elevated plus maze (Pellow et al. 1985). The social interaction test is thought to model generalized anxiety disorder and measures social interaction behaviors such as following, sniffing, and grooming as well as aggression toward a conspecific. Anxious rodents will tend to spend significantly less time in social interaction with a partner. The elevated plus maze is an elevated platform with four arms. Two opposing arms are “closed” with high arms, and the other two arms are open with no walls. The primary measures are percent time spent in open arms and number of open-arm entries. Most rodents will prefer spending exploring in the closed arms than the more threatening open arms. Decreased entries into and time spent in the open arms are considered a measure of anxiety. There can be two trials associated with the elevated plus maze. A first 5 min exposure is thought to model the escape components of panic disorder, whereas a second 5 min exposure is thought to model specific phobia (File et al. 2000).