Drugs and psychosis … and now for some good news

Whether at first contact or in ongoing care, around half of Australians with a diagnosis of psychosis also have a comorbid substance use disorder. In the current issue, Large and colleagues report a systematic meta-analysis which underlines the impact of ongoing substance use on psychotic symptoms (Large et al., 2014). They also touch on an intriguing finding: people with psychosis and substance disorders who cease drug use (‘former substance users’) may have better outcomes than people with psychosis who have never used substances. Put another way, substance use may be a positive prognostic sign in psychosis, as long as substance use is ceased.

There is growing evidence to support this claim. A meta-analysis of 23 studies by the same authors (Mullin et al., 2012) found that former substance users consistently have fewer positive symptoms and better functioning than people with psychosis who do not use substances. These are clinically significant effects; in more than 600 young people with early psychosis (Lambert et al., 2005), reducing substance use was the strongest predictor of remission. After controlling for age, sex, duration of untreated psychosis and symptom severity, decreasing or ceasing substance use more than doubled the likelihood of remission (hazard ratio 2.44). We have recently examined young people with a first psychosis admission in New South Wales. Of 4933 people who remained in contact with services over 2 years, those with ongoing substance problems had the highest readmission rate (66%), those with no substance disorders had an intermediate rate (50%) and those whose substance problems ceased had the lowest readmission rate (40%) (Sara et al., 2014b).

A link between substance use (even if discontinued) and positive outcome in psychosis seems counterintuitive. After all, substance use is associated with other factors usually linked to worse outcome in psychosis, including male gender, earlier onset of psychosis, social disadvantage and a family history of mental health and substance problems. Three explanations have been proposed to account for this apparent link between substance use and positive outcome: (i) direct chemical effects, (ii) social skills and (iii) personal vulnerability.

First, it is possible that substances themselves may have direct positive effects on brain function. Yucel and colleagues (2012) summarise evidence that cannabis use in people with psychosis is associated with better cognitive function as well as with fewer psychotic symptoms, leading to suggestions that cannabis may have direct neuro-protective effects or stimulate prefrontal neurotransmission. It is difficult to reconcile this theory with the harmful effects of ongoing cannabis use; however, our understanding of the effects of cannabis continues to evolve. For example, the interaction of THC and cannabidiols with endocannabinoid receptors is likely to involve both harmful and protective effects (Paparelli et al., 2011).

Second, Meuser et al. (1998) have proposed that the association of substances with positive outcome in psychosis is mediated through social competence, whereby more ‘socially oriented patients with serious mental illness are more likely to come into contact with drugs and subsequently develop substance use disorder’ (p.726). In this model, drug use is a marker for better premorbid functioning, greater social skill and therefore better outcome. Evidence to support this theory is limited. Prospective studies of drug use in adolescents typically find higher rates of drug use in more marginalised and distressed young people, not in those with greater social skills. Cannabis is ubiquitous in Australia and many other countries, requiring few social skills or networks to obtain. A ‘social skills’ theory would also predict that the association between drug use and cognition or outcome should be similar for different drug types and different diagnostic subtypes of psychosis, but there is evidence that these associations may differ for schizophrenia and bipolar disorder.

Third, these findings may reflect varying degrees of personal vulnerability. A stress-diathesis model suggests that onset of psychosis without substance use reflects a significant diathesis in the person affected, whereas substances may precipitate psychosis in individuals with less intrinsic vulnerability. Therefore, when those young people cease drugs, they have better outcomes. Consistent with this, people with psychosis and substance use have fewer neurological soft signs and fewer negative symptoms than those without substance abuse (Loberg and Hugdahl, 2009; Yucel et al., 2012). This theory would predict that substances with greater potential to precipitate psychotic symptoms, such as hallucinogens and stimulants, would trigger psychosis in people with lower levels of vulnerability. Hence, former users of those substances should have the most positive outcomes. A recent study found that drug-induced psychoses associated with amphetamines had a lower rate of transition to schizophrenia than those associated with cannabis (Niemi-Pynttari et al., 2013). We have also found that in people with brief, atypical and drug-induced psychoses, cannabis disorders predict a greater likelihood of transition to schizophrenia but stimulant disorders predict a lower likelihood (Sara et al., 2014a).

Regardless of the mechanisms involved, an association between ceasing substance use and positive outcome has important clinical implications. A first episode of a drug-associated psychosis is one of the most frightening crises that a young person and his or her family can face. By the time of that first presentation, risk factors such as family history, age and duration of untreated psychosis are fixed. Giving up substance use is one of the few ways through which a young person can modify his or her risk of developing a more enduring illness. If ceasing drug use not only helps avoid bad outcomes but can lead to good ones, this is an important and hopeful message. Two challenges for our mental health services follow. First, we need better ways of engaging and more effective interventions for people with comorbid substance use and psychosis. Second, more research is needed to identify whether the positive effects of ceasing substance use in people with early psychosis also apply for people with more enduring psychosis diagnoses such as schizophrenia.

See Review by Large et al., 2014, 48(5): 418–432.