Cannabis, Depression, and Bipolar Disorder

It is well-proven that cannabis use is associated with an increased risk of schizophrenia and heavy use of high-THC marijuana is associated with four times the normal rate of schizophrenia (see my previous posts, The Complex Link Between Cannabis Use and Psychosis and New Research Highlights the Cannabis Connection to Psychosis). But a connection between the affective disorders of unipolar depression and bipolar disorder has not been as clear. A sample of Swedish military conscripts with self-reported cannabis use that showed a dose-dependent associated risk of schizophrenia over time did not show a similar risk associated with unipolar depression or bipolar disorder. A new study involving the entire population of Denmark now provides interesting data associating a diagnosis of Cannabis Use Disorder (CUD) with subsequent diagnoses of affective disorders.

Since 1968, the Danish Civil Registration System has assigned a unique number to every permanent resident in the country. This number enables researchers to cross-reference groups between different registries. In a paper published May 24, Jefson[1], et. al., correlated the registry of individuals diagnosed with CUD in a treatment center or clinic with registries of individuals diagnosed at a later date with unipolar depression and bipolar disorder, including whether either diagnosis included psychotic features (Cannabis Use Disorder and Subsequent Risk of Psychotic and Nonpsychotic Unipolar Depression and Bipolar Disorder). The anonymity of individuals was protected. Since all diagnoses were dated, Jefson could ask whether a diagnosis of CUD was associated with later affective diagnoses at a rate differing from the rate seen in individuals without CUD. Any increased rates were expressed as hazard ratios (HR).

In this nationwide cohort study of over 6 million individuals, CUD was found to be associated with an increased subsequent risk of both unipolar depression and bipolar disorder diagnoses. Although excess risks of unipolar depression and bipolar disorder were highest immediately after diagnosis of CUD, they remained significantly elevated five to ten years after a diagnosis of CUD.

Individuals with CUD had a higher risk of unipolar depression (hazard ratio = 1.84) compared with individuals with no record of a CUD. No significant difference existed in the hazard ratios of psychotic versus nonpsychotic depression.

Of individuals with CUD, 14.1 percent eventually received a later diagnosis of bipolar disorder.

Most of these individuals (90.2 percent) were diagnosed with nonpsychotic bipolar disorder, while 9.8 percent were diagnosed with psychotic bipolar disorder. The hazard ratio of nonpsychotic bipolar disorder after a diagnosis of CUD was higher than for unipolar depression; 2.96 in men and 2.60 in women. The hazard ratio for developing psychotic bipolar disorder rose to 4.05. The jump in risk for psychotic bipolar disorder over nonpsychotic bipolar disorder, when combined with studies establishing the risk of schizophrenia after cannabis use, has led to speculation that cannabis has a psychotogenic effect independent from its impact on affect.

Few studies have as massive a sample of individuals being studied as Jefson’s n of over 6 million. But what does this study really tell us? From the outset, Jefson only looked at diagnosed cases of CUD. Far more Danes could have been using cannabis regularly without being diagnosed. If they had also been counted, the hazard ratio for bipolar disorder might have been significantly lower. This is assuming that undiagnosed CUD would be more common than undiagnosed bipolar disorder. Undiagnosed depression, of course, might even be more common than undiagnosed CUD. Furthermore, a diagnosis of CUD in a treatment center or clinic may have alerted clinicians to look more closely for associated comorbidities such as depression and bipolar disorder.

On the other hand, this massive study adds additional weight to the possibility of cannabis possessing psychotogenic power in vulnerable individuals. At the least, the study should alert all clinicians suspecting a patient of using cannabis too liberally to also look for subsequent affective disorders. While they are not common, this Danish survey shows that neither are they rare.