Co-Occurring Bipolar and Substance Use Disorders: A Review of Impacts, Biopsychosocial Mechanisms, Assessment, and Treatment

Abstract

Individuals with bipolar disorder are at significant risk for having a co-occurring substance use disorder—particularly, alcohol and cannabis use disorders. Having a co-occurring substance use disorder is associated with a more pernicious clinical course, lower quality of life, and poorer treatment outcomes. Despite its increased morbidity, there is little research and clinical evidence-based guidelines on the treatment of individuals with co-occurring bipolar and substance use disorders. This review details current knowledge on the prevalence, clinical correlates, and biopsychosocial mechanisms underlying co-occurring bipolar and substance use disorders. The authors present recent research that highlights underlying mechanisms of comorbidity, including aberrant reward processing, stress sensitization, early childhood maltreatment, and gene–environment interactions. Next, the authors review current evidence-based recommendations for the assessment and treatment of co-occurring bipolar and substance use disorder, highlighting areas of needed future clinical research.

Individuals with bipolar disorder often experience co-occurring substance use disorders at significantly higher rates than the general population (1). In a large epidemiological study (2), individuals with bipolar disorder had a lifetime prevalence of any substance use disorder of 56.1% (60.7% for bipolar I disorder and 48.1% for bipolar II disorder), compared with 13.8% of the general population. A more recent study (3) that combined longitudinal research cohorts of individuals with bipolar disorder (N=3,980) reported an overall prevalence rate of any substance use disorder to be 54.9%, which indicates that these rates appear to be stable across time.

The two most reported substance use disorders among individuals with bipolar disorder include alcohol use disorder and cannabis use disorder, followed by substance use disorders involving cocaine, stimulants, and other drugs (2, 4). Across 78 clinical studies, alcohol use disorder was present in an average of 30% of individuals with bipolar disorder, with the highest rate across studies reported at 67% (5). In a recent meta-analysis and review, the estimated prevalence of cannabis use disorder among individuals with bipolar disorder was 24% across 53 studies, with the highest prevalence reported at 67% (6). In clinical studies, the lifetime prevalence of cannabis use disorder is reported to be 36% (7). The prevalence of other substance use disorders, primarily cocaine and stimulant use disorders, is estimated to be 17% among individuals with bipolar disorder (8). Of note, it is reported that 60%–70% of individuals with bipolar disorder have a nicotine or tobacco use disorder; however, nicotine and tobacco often do not get included in studies of “any substance use disorder” such as those mentioned earlier (9), which indicates that the overall rates of comorbidity may be even higher.

In terms of important demographic correlates of co-occurring bipolar and substance use disorders, men tend to use substances at higher rates than women, and use appears to decrease as these individuals age (7). Although prevalence rates are higher among men, the relative risk of having a co-occurring substance use disorder is higher among women when considering rates of substance use disorders among women in the general population (5, 10). In terms of bipolar disorder subtypes, some research suggests that individuals with bipolar I disorder have higher rates of substance use disorders compared with those with bipolar II disorder (11); however, the impact of having a co-occurring substance use disorder may be more profound for those with bipolar II disorder (12).

Impact on Presentation, Symptoms, and Functioning

Overall, co-occurring bipolar and substance use disorders are associated with a more pernicious illness course (10, 13), with mood symptoms that are more frequent, more severe, and of longer duration (14–18). In addition, overall, co-occurring bipolar and substance use disorders are related to greater impulsivity (19), greater cognitive deficits (20, 21), and an even greater risk for suicide-related thoughts and behaviors (22–24), compared with when these disorders are present on their own. Co-occurring alcohol use disorder and bipolar disorder are specifically associated with decreased functioning, poorer recovery after mood episodes, more severe mood symptoms, and increased suicidality and mortality (25–27). In a 2024 study, even small increases in alcohol use were associated with increases in depressive and manic symptoms over the following 6 months in individuals with bipolar I or II disorder (12). Similarly, bipolar disorder with co-occurring cannabis use disorder has been associated with less time in remission from depressive and manic episodes (28), more severe depressive and manic symptoms (29), and poorer functioning and quality of life (30).

Substance use disorders may unmask mood symptoms in vulnerable individuals who, without substance use, would have never developed or later developed bipolar disorder (16, 31, 32). For example, alcohol generates reactive oxidative species, highly reactive molecules that contain oxygen, within brain regions that have been implicated in bipolar disorder (33, 34). Recent work suggests that young adults with bipolar disorder, and those who are at risk for bipolar disorder, may show distinct neural correlates of alcohol use (35). As increased oxidative stress in heavy drinkers with bipolar disorder is reported (36), neurotoxic effects of alcohol may be one mechanism that contributes to clinical progression among individuals with bipolar disorder (37). Indeed, with increasing withdrawal episodes, it is suggested that neurons become more likely to depolarize autonomously, which may trigger mood episodes in the absence of stressors (38).

Although these studies illustrate one pathway to addiction (i.e., substance use that affects illness course or contributes to the onset of mood symptoms), it is important to emphasize that the relation between these disorders are bidirectional, with mood symptoms and mood disorders affecting substance use and the likelihood of developing problems over time. This latter path to comorbidity has received less research attention, but work is emerging that suggests an increased sensitivity to substance use and unique risk factors that may drive the risk and onset of substance use disorders among individuals with bipolar disorder. Collectively, these findings highlight complex interactions between bipolar and substance use disorders, as well as the critical need for more research on the mechanisms underlying development, maintenance, and treatment of co-occurring bipolar and substance use disorders.

Mechanisms Underlying Co-Occurrence

Going Beyond the Self-Medication Hypothesis

It is important to note that a self-medication hypothesis is one model proposed to explain the high rates of comorbidity between bipolar disorder and substance use disorders. Although drinking to cope has been endorsed among individuals with bipolar disorder (39), and some (but not all) studies suggest greater alcohol and drug use during acute mood episodes (17, 40), the self-medication hypothesis cannot unequivocally explain the frequent co-occurrence between these conditions (41). Recent studies are emerging to suggest novel risk factors that may drive development of this comorbidity, independent of mood symptoms. For example, it was recently reported that youths with bipolar disorder (diagnosed before the development of alcohol and/or substance use disorders) show differences in subjective intoxication and neural responses to alcohol (42, 43), with these differences related to prospective alcohol use (44). Differences in subjective experience of intoxication is an established risk factor for alcohol use disorder (45–47). Greater stress sensitization is also suggested to contribute to risk for bipolar and substance use disorders (48, 49). These studies are examples meant to highlight risk factors that extend beyond the self-medication hypothesis to contribute to comorbidity. Our understanding of risk factors and development of this comorbidity is sparse. Although few studies have investigated comorbid conditions in youths with substance use disorders, there is evidence that over 60% of youths with substance use disorders (identified in community-based treatment programs for substance use disorders) also meet the diagnostic criteria for a comorbid mental illness (50). Comorbidity is, therefore, the norm and not the exception, yet most studies that have investigated the risk factors for substance use disorders exclude individuals with mental illness, especially those with bipolar disorder. This greatly limits generalizability. A greater understanding is needed regarding mechanisms that contribute to the risk and development of substance use disorders in individuals with bipolar disorder to inform early interventions and foster prevention efforts that are more specific to this vulnerable population. Risk factors may include, but are not limited to, early life stress and genetic vulnerability, and they likely relate to gene–environment interactions as briefly highlighted in the following text.

Early Life Stress

Early life stress is a well-established risk factor for bipolar disorder, substance use disorders, and their co-occurrence. An abundance of literature has reported that childhood maltreatment is associated with adolescent substance use, the onset of mood symptoms and bipolar disorder, and a worse bipolar illness course (51–53). Adolescent substance use and mood-related symptoms, in turn, increase the risk for adult mood and substance use disorders (54, 55). Early life stress is suggested to be even more prevalent in individuals who present with co-occurring bipolar and substance use disorders. For example, in a study conducted by Wu and colleagues (2010), 95% of individuals with mental health conditions and co-occurring substance use disorders reported childhood trauma (56). Although the effects of early life stress could be a general shared risk factor for psychiatric outcomes and not specific to mood and substance use disorders, it has been suggested that mood disorders may precede substance misuse in adolescents after early life stress (57–59). For example, in a study of 690 adults, depression mediated the effect of childhood maltreatment on later alcohol-related problems (60). Childhood maltreatment is associated with longer and more severe withdrawal symptoms during detoxification (61), and depression severity has also been observed to relate to the severity of withdrawal symptoms during detoxification among individuals who report early life stress (62).

Most studies that have investigated the intersection of mood disorder, substance use disorder, and early life stress have focused on depression symptoms in general and have not examined bipolar disorder specifically, often excluding those diagnosed as having bipolar disorder. However, a history of childhood maltreatment is associated with higher risk for comorbid substance use disorders among individuals with bipolar disorder, compared with individuals with bipolar disorder without a history of childhood maltreatment (63–65). Childhood maltreatment is associated with greater impulsivity in individuals with bipolar disorder (66). Although developmental studies investigating variation in trait impulsivity in bipolar disorder and relations with alcohol use or abuse have not suggested an association (44, 67), differences in risk-taking behavior—specifically, greater trait impulsivity and loss aversion—are reported among individuals with bipolar disorder and a current (19) as well as prior history of substance use disorders (68), respectively. Additionally, childhood maltreatment is associated with positive and negative urgency (69), a tendency to act impulsively when experiencing positive or negative emotions, respectively. Positive and negative urgency are suggested to mediate the relations between early life stress and substance use (70–72). A recent study by Brand et al. (2024) observed negative urgency to be higher in individuals with co-occurring bipolar and alcohol use disorder compared with individuals with only bipolar disorder or alcohol use disorder (73). Negative urgency related to rapid cycling only in individuals with comorbidity (not individuals with bipolar disorder alone) and total scores on the Alcohol Dependence Scale related to positive and negative urgency subscales only in individuals with comorbidity (not individuals with alcohol use disorder alone). Collectively, results could converge to suggest emotional impulsivity as an area of vulnerability and facets of impulsivity and risk taking—even during remission—that could contribute to the maintenance and relapse of alcohol/substance use in bipolar disorder.

More research is needed to further our understanding of mechanisms that contribute to substance use disorders in individuals with bipolar disorder, including investigating the role(s) of early life stress across the addiction cycle (i.e., substance misuse, disorder onset, maintenance, and relapse). Given that the extant literature is largely retrospective in nature, prospective longitudinal studies are critical for establishing causality. Furthermore, as not everyone with early life stress will develop bipolar disorder and/or a substance use disorder, genetic factors likely increase sensitivity to early life stress and relate to psychiatric outcomes.

Genetic Vulnerability

Given the familial nature of both bipolar disorder and substance use disorders, environmental and genetic mechanisms likely contribute to the emergence of comorbidity. Genes that confer risk for bipolar disorder may also increase risk for substance use disorders, as there is overlap in the genes implicated in both conditions (74, 75). Although family studies investigating the co-occurrence of bipolar and substance use disorders have produced mixed results, it is suggested that adolescents with bipolar disorder have an increased prevalence of substance use disorders in relatives (76). However, there is also a greater risk for co-occurring substance use disorders in offspring with bipolar disorder, even without a family history of substance use disorders (77, 78), although an additive effect has been suggested (79). These studies indicate complex familial etiologic risk factors. Additionally, genetic vulnerability–environment interactions are suggested to contribute to psychiatric outcomes, including bipolar and substance use disorders. In a recent placebo-controlled alcohol administration study by Kosted et al. (2023) (42), interactions between early life stress (measured with the Childhood Trauma Questionnaire) and parental risk for mood (including unipolar depression and bipolar disorder) and/or substance use problems was observed on subjective experience of intoxication. Specifically, greater reported childhood trauma related to feeling more drunk during beverage sessions in youths with parental risk for bipolar disorder and/or substance use problems, but not youths without parental risk. The effect of early life stress was also more pronounced in individuals with parental risk for mood and/or substance use disorders if the drinking occurred in a simulated bar (42). This finding could suggest a role of early life stress in individuals with familial risk for mood and/or substance use disorders in alcohol cue reactivity or alcohol expectancies that could serve as targets for early interventions. Indeed, childhood maltreatment–related changes in paralimbic prefrontal systems have been observed to relate to substance use in individuals with familial risk for bipolar disorder (35), and paralimbic prefrontal systems are associated with alcohol expectancies (80, 81) and alcohol cue reactivity (82). A recent study of adults with co-occurring bipolar and alcohol use disorder showed differences in paralimbic prefrontal system reactivity (specifically, hypoactivation in insula and inferior frontal gyrus) during alcohol cue processing compared with adults with bipolar or alcohol use disorder alone, with these differences relating to alcohol craving exclusively among individuals with co-occurring bipolar and alcohol use disorder (83). Although reviewing neural systems that are suggested to underlie this comorbidity is beyond the scope of this review, findings converge to suggest gene–environment interactions on neural phenotypes and behavioral outcomes related to the risk and maintenance of comorbid substance use disorders in bipolar disorder.

Mechanisms at the Molecular Level

There is an overlap in molecular mechanisms that are implicated in bipolar disorder as well as substance use disorders. For example, alterations in the N-methyl-d-aspartate (NMDA) receptor are supported in bipolar disorder (84) and thought to confer risk (85) and be a promising target for interventions (86). Acute alcohol use affects NMDA dynamics (87, 88), and glutamatergic systems are thought to contribute to the risk for alcohol use disorder as well as other behavioral effects of alcohol, such as withdrawal symptoms (89). Targeting NMDA dynamics could be an optimal treatment strategy for treating co-occurring bipolar and alcohol use disorder. Indeed, when following 27 patients with bipolar disorder, Chitty et al. (2015) observed mismatched negativity—a marker of NMDA function—to relate to changes in alcohol use over time (90). Pharmacotherapies that target NMDA receptors (e.g., aripiprazole) are also suggested to decrease craving in individuals with comorbidity (91). Other overlapping molecular mechanisms implicated in both bipolar disorder and substance use disorders include, but are not limited to, serotonin receptor subtypes, corticotropin-releasing hormone receptor 1 gene, glucocorticoid receptors, inflammation, and the hypothalamus–anterior pituitary–adrenal axis (53, 54). A detailed examination of mechanisms at the molecular level that may be involved in the risk for—and development and maintenance of—co-occurring bipolar and substance use disorders is beyond the scope of this review; however, it is important to emphasize that more research in this area is needed. A better understanding of molecular mechanisms in bipolar disorder that may interact with substances could inform novel pharmacotherapies in better treatment of comorbidity, as previously discussed (92).

Clinical Assessment Considerations

Given the clear impact that substance use has on bipolar disorder symptoms, functioning, and treatment response, as well as the impact of mood on substance use, it is critical that proper assessment of substance use patterns is a part of standard care for individuals with bipolar disorder. Often, the goal of assessment is to rule out a substance use disorder; however, emerging research suggests that even moderate levels of substance use and small increases in alcohol use is associated with worse illness outcomes in individuals with bipolar disorder (12, 15). Therefore, we recommend dimensional measures that assess the intensity, frequency, and impairment related to the use of different substances—measures that provide clear thresholds for probable use disorders but that also allow for the assessment of within-person changes in use over the course of illness or treatment. In the following text, we outline measures (Table 1) that have a strong evidence base (93) and/or are endorsed by the National Institute on Alcohol Abuse and Alcoholism, the National Institute on Drug Abuse, or the World Health Organization. More extensive details regarding the assessment of substance use in individuals with bipolar disorder are available elsewhere (94).

TABLE 1.

Recommended measures for assessing substance use in bipolar disorder^a

Substance	Measure name	Clinician interview	Self-administered	Scoring and thresholds
Alcohol	Alcohol Use Disorder Identification Test (AUDIT)	✓	✓	AUDIT (10-item version) score range: 0–40; 0–7 = low-risk drinking; 8–14 = potential harmful or hazardous drinking; ≥15 = high likelihood of moderate to severe alcohol use disorder
​	​	​	​	AUDIT-C (brief three-item version) score range: 0–12; ≥4 = positive screen for men; ≥3 = positive screen for women;
Tobacco, alcohol, cannabis, cocaine, stimulants, inhalants, sedatives, hallucinogens, opioids, other drugs	Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST)	✓	​	Cannabis and other drugs: 0–3 = low risk; 4–26 = moderate risk; ≥27 = high risk Alcohol: 0–10 = low risk; 11–26 = moderate risk; ≥27 = high risk
Cannabis	Cannabis Use Disorder Identification Test—Revised (CUDIT–R)	​	✓	CUDIT–R score range: 0–32; 0–7 = low-risk cannabis use; 8–11 = hazardous cannabis use; ≥12 = high likelihood of cannabis use disorder
Cannabis, amphetamines, cocaine, opiates, hallucinogens, solvents–inhalants, GHB, prescription pills and other substances	Drug Use Disorder Identification Test (DUDIT)	​	✓	DUDIT score range: 0–44; ≥6 = general population screen cutoff for males; ≥2 = general population screen cutoff for female patients; ≥25 = likely substance use disorder with physiological dependence
Cannabis, solvents, tranquilizers, barbituates, cocaine, stimulants	Ten-item Drug Abuse Screening Test (DAST-10)	✓	✓	DAST-10 scores: 1–2 = low level of problems, recommendation for monitoring and reassessment; 3–5 = moderate level of problems that warrants further assessment; 6–8 = substantial level of problems that warrants intensive assessment; 9–10 = severe level of problems that warrant intensive assessment
Tobacco, alcohol, prescription medications, other substances	Tobacco, Alcohol, Prescription Medications, and Other Substances tool (TAPS; TAPS-2)	✓	✓	If patients endorse use of any substances, administer the TAPS-2. Scoring of TAPS-2: 0 = no use in the past 3 months; 1 = problem use; 2 = high-risk use
Alcohol use, smoking, other	Patient-Reported Outcomes Measurement Information System (PROMIS)	​	✓	Unlike other measures, these are multidimensional, and you can “pick and choose” subscales relevant to treatment. These substances do not have typical cutoff scores like other measures, so they are best suited for measuring change in treatment.
​	​	​	​	Alcohol: severity of alcohol use; negative consequences of alcohol use; negative expectancies related to alcohol use; positive consequences of alcohol use; positive expectancies related to alcohol use
​	​	​	​	Smoking: coping expectancies; emotional-sensory expectancies; negative health expectancies; negative psychosocial expectancies; nicotine dependence; social motivations
​	​	​	​	Substance use: appeal of substance use; prescription medication misuse; severity of substance use

^a GHB, gamma-hydroxybutyric acid; TAPS-2, a brief assessment adapted from the ASSIST-Lite.

Note that individuals with bipolar disorder may have different motivations for using substances (e.g., alcohol to sleep, cocaine or other stimulants to energize, cannabis to relax) that can be critical to assess in the context of treatment for co-occurring bipolar and substance use disorders. As such, we do recommend implementing measures from the Patient-Reported Outcomes Measurement Information System (or, PROMIS; detailed in Table 1) or using motivational interviewing to understand the motivations and expectancies for use as well as the fears and ambivalence about changing one’s substance use behavior. In addition, polysubstance use is a common presentation among individuals with bipolar disorder, which brings additional challenges to assessment. This is particularly true in the case of alcohol and cannabis. As such, utilizing assessment measures that capture multiple substance use patterns can be helpful.

Treatment Considerations for Co-Occurring Bipolar and Substance Use Disorders

Individuals with co-occurring bipolar and substance use disorders often have greater treatment resistance and relapse risk. Although this may make clinical intervention difficult, it also highlights the critical need for increasing evidence-based intervention options for this population. A recent review of 24 guidelines concluded that integrated treatments that focus on both psychiatric disorders and substance use in the same setting is the optimal approach (95). However, alternative approaches include a parallel design, in which distinct services are being provided at the same time, or a sequential design, in which the “primary” disorder is treated first have been implemented.

Older work has suggested that targeting the primary disorder may be an optimal treatment strategy in individuals with comorbidity (96). Patients who present with a primary mood disorder (mood disorder onset first) may be a distinct population compared with patients who present with a primary substance use disorder (16, 31, 97). Determining the temporal relations between onset of these disorders may be important in fostering the most effective treatment plan (96). For example, a recent investigation of youths with co-occurring bipolar and substance use disorders reported that, in adolescents with primary bipolar disorder, lithium treatment improved outcomes and reduced the number of positive urine screens for recreational substance use (98). Targeting the primary disorder may be beneficial, but there is little work in this area and identifying which disorder is primary can be burdensome, if not impossible, in some complex clinical presentations. Although they are, perhaps, the most common approach, “sequential” designs have the least evidence (95).

Research focused on integrated and parallel design approaches in treating bipolar disorder have largely focused on psychotherapy. A systematic review identified five studies on psychological therapy for co-occurring bipolar and substance use disorders (99), with four of five studies reporting a positive impact of psychotherapy. Two studies found that although integrative therapy reduced both mood and substance use symptoms, they did not do so more than psychotherapy for bipolar disorder alone. A more recent meta-analysis reviewed eight trials of co-occurring bipolar and substance use, including those focused on smoking cessation (100). They concluded that there was little evidence of an impact on both mood and substance use symptoms together, with integrated group therapy showing improved substance use outcomes for bipolar disorder and other interventions showing improved mood symptoms but no impact on substance use. Nonetheless, the Canadian Network for Mood and Anxiety Treatments Task Force (CANMAT) (101), the most current consensus guidelines for the treatment of co-occurring bipolar and substance use disorders, reports that the best psychotherapy evidence is integrated group therapy, which combines a cognitive-behavioral therapy approach focused on relapse prevention of both substance use and mood episodes, maladaptive cognitions that underlie both, and some session-specific content on substance use and bipolar disorder (102). Less evidence supports the utilization of motivational interviewing alone or adding it to cognitive-behavioral therapy. Although the evidence is mixed, at present, there is more evidence that integrated psychotherapies are ideal for the treatment of substance use disorders in individuals with bipolar disorder over traditional parallel (distinct treatments at once) or sequential (one treated, then the other) treatment approaches.

It is possible that traditional evidence-based therapies for bipolar disorder, which are not developed to target substance use, could have beneficial impacts on substance use in individuals with bipolar disorder. Unfortunately, our ability to test this is hampered by the fact that most randomized clinical trials exclude individuals with bipolar disorder who have current alcohol or substance use. One preliminary study that allowed for current and lifetime substance use showed that social and functional improvements after interpersonal and social rhythms therapy—an evidence-based therapy for bipolar disorder—were equally as strong for those with a lifetime or current substance use disorder as for those without it (103); however, the researchers did not study the impact on substance use, as this was a secondary analysis and did not include such outcome measures. There is a critical need for clinical trials of existing and novel psychotherapy to integrate individuals with bipolar disorder and co-occurring substance use.

In terms of psychopharmacology, certain medications that are commonly prescribed for bipolar disorder are suggested to contribute to differences in alcohol and substance use. Although exploratory, a recent placebo-controlled alcohol administration study in young adults with bipolar disorder reported a main effect of anticonvulsant use on subjective response to alcohol, with those who took an anticonvulsant feeling more “woozy” or “dizzy” during beverage sessions (104). Sperry et al. (2024) recently completed a study of 584 individuals with bipolar disorder that examined relations between alcohol use and bipolar disorder over time (12). Benzodiazepine and antipsychotic use related to lower longitudinal levels of alcohol use. It is important to note that conflicting data were recently reported from an electronic medical record analysis (105) suggesting that benzodiazepine use may positively relate to substance use. Physicians do not prescribe at random, and emerging problems with substance use may confound treatment decision making in this population. Although the effects of medication should be interpreted with caution and serve as a proof of concept for future studies, our findings highlight the need for more randomized clinical trials to disentangle the risks and benefits of medication on substance use in bipolar disorder.

The CANMAT guidelines, although over a decade old, are still the current gold-standard consensus guidelines for the psychopharmacological treatment of co-occurring bipolar and substance use disorders. A detailed description of their findings is not provided, as guidelines are widely available; however, we highlight here the guidelines that met the criteria for level 2 evidence (none met the criteria for level 1). Mood stabilizers, including lithium and valproic acid, are recommended, with level 2 evidence for co-occurring alcohol, cannabis, and cocaine use. It is important to note that valproic acid should not be used by women of childbearing age (106). Antipsychotics, including quetiapine, have level 2 evidence for co-occurring cocaine, amphetamine, and methamphetamine use. Risperidone has level 2 evidence for cocaine and methamphetamine use. Other agents, such as citicoline, have level 2 evidence for co-occurring cocaine use, disulfiram for co-occurring alcohol use, and naltrexone for co-occurring alcohol use. Note that since these guidelines were published, disulfiram has been removed from the market in some areas because of rare but deadly side effects when combined with alcohol (107). Although acamprosate has been widely used for the treatment of alcohol use disorder, negative results have been reported in studies of individuals with bipolar disorder and co-occurring alcohol use disorder (108). Recent research shows some potential for other pharmacological agents such as gabapentin for the treatment of co-occurring cannabis use disorder and bipolar disorder; however, this was a small trial and should be interpreted with caution (109). Furthermore, a recent review highlighted several newer pharmacological agents that may have potential for co-occurring bipolar and alcohol use disorder, including baclofen, varenicline, gabapentin, and topiramate; however, formal recommendations have not yet been made (107). Additionally, low-dose ketamine has shown some potential for alcohol use disorder (110) and bipolar disorders (111) (separately), with its mechanism of action proposed through reducing maladaptive reward memories and improved psychotherapy adherence. However, these data are preliminary, and there could be unintended side effects of the use of ketamine (e.g., psychosis) (112). This recent review highlights that the current best guidelines for bipolar disorder and alcohol use disorder stabilization remain true from the CANMAT guidelines: lithium; anticonvulsants, such as valproate; and antipsychotics, with naltrexone as an adjunctive for reducing craving.

Future treatment research could also focus on interventions that target the mechanisms underlying the primary risk factors that contribute to bipolar disorder, substance use disorders, and their co-occurrence. For example, as discussed earlier, early life stress is a primary risk factor for both bipolar disorder, substance use disorders, and their co-occurrence. Systemic inflammation is associated with early life stress and is a risk factor for bipolar and substance use disorders (113, 114). There is a paucity of studies that investigate the degree to which early life stress–related inflammation contributes to co-occurring bipolar disorder and substance use disorders. Given the roles of inflammation in both conditions, inflammation is a likely targetable mechanism that contributes to the frequent comorbidity between these disorders (115). Although preliminary, clinical trials that investigate anti-inflammatory therapies in individuals with bipolar disorder and early life stress suggest that inflammation could be a promising therapeutic target (53). As early life stress is an established risk factor for co-occurrence between bipolar and substance use disorders (reviewed earlier), and clinical trials have suggested benefits of anti-inflammatory therapies in individuals with substance use disorders (116, 117), more research is needed that investigates inflammation-related mechanisms that contribute to the risk, onset, and progression of comorbidity and whether targeting could improve outcomes in this population.

Conclusions

Despite the high prevalence of co-occurring bipolar and substance use disorders as well as increased morbidity and mortality, clinical research focused on this co-occurrence is in its infancy. However, exciting recent research highlights important etiological areas for target identification and target engagement, including early childhood maltreatment, neural abnormalities in reward systems, emotion-related impulsivity, and gene–environment interactions. Although research is lagging, proper assessment and intervention is critical for individuals with this comorbidity. We provide recommendations for the use of evidence-based assessment measures in the standard care of individuals with bipolar disorder, and we review current best practices for the psychological and pharmacological treatment of individuals with co-occurring bipolar and substance use disorders. A review of the extant literature highlights that much existing work on pharmacological interventions for co-occurring bipolar and alcohol use disorder is of low grade and that the development of novel agents with more focused targets on shared etiological factors is warranted.