The findings were published online November 28 in the American Journal of Psychiatry.
“Substance-induced psychoses are interesting to investigate since we know very little about them ― and less than you might think,” said Dr Hjorthøj.
“Moreover, the potential etiology between substance use and later development of severe mental illness is poorly understood; and it is a continuing point of debate whether the connection is at all causal. The present study is a potentially important link in this debate.”
He added that because the incidence of substance-induced psychoses, especially from cannabis use, has been increasing, it’s important to determine whether there are long-term effects.
“This could arguably be even more important in a world in which we see an increasing tendency towards legalization or decriminalization of cannabis use,” said Dr Hjorthøj.
Also, “a better understanding of what other factors may drive this association will help us to figure out what to do, in terms of targeted prevention and tailored follow-ups.”
The investigators examined data from the Danish Civil Registration System and the Psychiatric Central Research Register for 6788 patients (75% men) who were diagnosed between 1994 and 2014 with substance-induced psychosis. Patients with a prior diagnosis of BD or schizophrenia spectrum disorders were excluded from the study.
Significant Conversion Rates
Alcohol-induced psychosis was the most common diagnosis (n = 1680), followed by psychosis induced by mixed/other substances (n = 1405) and by cannabis (n = 1222).
Interestingly, the mean age of onset of psychosis from alcohol, opioids, or sedatives was older than 45 years. It was 30 years or younger for cannabis, amphetamines, hallucinogens, cocaine, and “other substances.”
After any substance-induced psychosis, the 20 year-conversion rates were as follows:
- 32.2% for either schizophrenia or BD (95% confidence interval [CI], 29.7 – 34.9)
- 26.0% for just schizophrenia (95% CI, 23.7 0 28.9)
- 8.4% for just BD (95% CI, 7.4 – 9.5)
The highest rate of conversion to either schizophrenia or BD or to schizophrenia alone was from cannabis-induced psychosis (47.4% and 41.2%, respectively).
The following table shows conversion rates for other types of substances.
Table. Rates of Conversion to Schizophrenia or BD After Substance-Induced Psychosis
||Conversion Rate (%)
||31.8 – 38.3
||26.0 – 39.7
||19.5 – 38.6
||17.6 – 27.5
||11.9 – 35.1
||13.7 – 29.3
||12.8 – 30.1
The investigators also matched these patients to a healthy comparator group. The hazard ratio (HR) was 73.3 for conversion to schizophrenia in those with substance-induced psychosis vs the comparator group (95% CI, 65.2 – 91.7; P < .001). It was 24.4 for conversion to BD (95% CI, 20.1 – 29.6; P < .001).
The substance associated with the highest risk for conversion to schizophrenia was cannabis (HR, 101.7; 95% CI, 74.1 – 139.7; P < .001). Its HR for conversion to BD was 32.5 (95% CI, 21.1 – 50.0; P < .001).
All types of substance-induced psychosis except for psychosis related to sedatives were significantly associated with increased risk for schizophrenia and for BD (all comparisons, P < .001). Sedative-induced psychosis was associated with conversion to BD only (HR, 33.0).
“We found that 50% of conversions to schizophrenia occurred within 3.1 years [after substance-induced psychosis], and 50% of conversions to bipolar disorder occurred within 4.4 years,” write the investigators.
Among the subgroup who experienced cannabis-induced psychosis, half the men converted to schizophrenia within 2 years, and half the women converted within 4.4 years.
The risk for schizophrenia decreased as age at time of psychosis induced by any substance increased. The highest risk was for those between the ages of 16 and 25. This pattern was not found for conversion to BD.
The HR was 1.92 for conversion to schizophrenia for self-harm after the psychosis episode (95% CI, 1.58 – 2.34); it was 1.60 for conversion to BD (95% CI, 1.13 – 2.27).
Need for Long-term Follow-up
“Based on the different risk factors identified in different analyses and the overall conversion rate of 32.2%, it seems most reasonable to suggest that all patients with a substance-induced psychosis should be offered follow-up,” write the investigators.
They add that this follow-up period should be at least 2 years, but that longer follow-up may prevent incidents of self-harm. “In Denmark, patients are offered follow-up after self-harm, but not after a substance-induced psychosis,” they write.
In the study, “the magnitude of the problem of conversions was somewhat surprising,” added Dr Hjorthøj. “Substance-induced psychosis is clearly not harmless when nearly a third of all cases later develop schizophrenia or bipolar disorder.”
He noted that he was surprised to not find any groups that were at decidedly low risk for conversion, “which is why we conclude that all patients with substance-induced psychosis should be monitored closely.”
Dr Hjorthøj and one of the two other study authors have disclosed no relevant financial relationships. The other author received a pregraduate research grant from the Lundbeck Foundation for work reported in the article.