Tobacco History Complicates Cannabis and Psychosis Research

When researchers examine cannabis and mental health, tobacco can look like a secondary detail. A new study suggests that it deserves a more central role. Among people with psychotic disorders, the relationship between cannabis use and psychiatric symptoms differed depending on whether participants currently smoked tobacco or had never smoked it.

The study, published in Schizophrenia Research, analyzed 884 people with schizophrenia, schizoaffective disorder, or bipolar disorder with psychotic features.¹ Its findings offer a useful direction for cannabis research: understanding who uses cannabis, what else they use, and how those patterns relate to their health.

The opportunity is greater precision. Better distinctions between substance use patterns could help researchers interpret conflicting findings and help clinicians ask more informative questions. That progress does not require presenting cannabis as either a universal threat or a treatment for psychosis.

What the Cannabis and Tobacco Study Examined

Researchers used data from the Bipolar-Schizophrenia Network on Intermediate Phenotypes 2, a multisite project that collected clinical and cognitive assessments between 2015 and 2020. Participants were taking stable medication, with no changes in medication or dose during the preceding 30 days.

The team compared five groups based on substance use during the previous month. These included people reporting no substance use, people using substances other than cannabis or tobacco, tobacco-only users, cannabis-only users, and people using both cannabis and tobacco.

Those labels require care. “Tobacco-only” meant no recent cannabis use, but did not necessarily mean no alcohol or other substances. Likewise, the cannabis-only and co-use groups could report other substance use. This was a comparison of tobacco and cannabis patterns within a broader clinical population.

The researchers also examined tobacco history alongside an adult cannabis exposure measure. Participants reported how many times they had used cannabis after age 18, using categories ranging from none to approximately 100 or more occasions. This measures accumulated experience, rather than grams consumed, THC strength, or a current daily schedule.

How Tobacco History Changed the Findings

Tobacco-only users had greater psychiatric symptom burden than participants using neither tobacco nor cannabis, and poorer cognitive performance than several comparison groups. Tobacco use therefore appeared to identify a subgroup with more substantial difficulties, although the study could not determine why.

The cannabis findings were more complicated. Greater adult cannabis use was associated with worse general psychiatric symptoms among current tobacco users, but fewer general symptoms among people who had never used tobacco.

“General symptoms” is an important distinction. The analysis does not mean cannabis reduced hallucinations or delusions. The general psychopathology measure is separate from the scales assessing positive and negative symptoms of psychosis. Collapsing these outcomes into a claim that cannabis “improves psychosis” would misrepresent the results.

Why Better Cognitive Scores Do Not Establish a Benefit

Greater adult cannabis use was also associated with better global cognition, processing speed, and working memory among current tobacco users. The working-memory relationship differed between current and past tobacco users, whereas the other cognitive relationships did not show significant differences by tobacco history.

The authors offer a possible explanation: people with better underlying cognitive functioning may have been more able to obtain and use cannabis. Recreational cannabis was illegal at most study sites during data collection, making access a potentially relevant factor.

Imagine two participants with the same diagnosis but different abilities to organize daily activities and maintain social connections. The person who can more easily arrange purchases may report more cannabis use and perform better on cognitive testing. Cannabis would then be associated with better performance without having produced it. The study cannot distinguish that explanation from other possibilities.

Psychosis Risk and Existing Symptoms Are Different Questions

This research concerns people who already had psychotic disorders. It does not establish the risk that cannabis use will lead someone without such a diagnosis to develop one. Keeping those questions separate allows findings from different populations to inform each other without being treated as interchangeable.

A separate 2026 study in Nature Mental Health examined cannabis and tobacco co-use in people at clinical high risk for psychosis. Heavy cannabis use combined with light tobacco use was associated with a higher risk of developing psychosis than using neither substance. However, the simpler comparison of any co-use versus neither substance did not reach statistical significance.

That distinction matters. Even within one study, a broad yes-or-no category can produce a different result from an analysis that considers exposure intensity. Neither study establishes causality, but both demonstrate why tobacco deserves more attention when interpreting cannabis findings.

Another 2026 study of co-use and cognitive performance found lower scores on several measures among high-risk co-users compared with healthy controls. High-risk participants reporting no substance use also showed cognitive difficulties. The comparison group matters as much as the headline: a difference from healthy controls cannot by itself identify a substance-specific effect.

Better Cannabis Research Starts With Better Measurements

The practical lesson extends beyond tobacco. Cannabis exposure is not a single, uniform experience. A lifetime category of more than 100 occasions cannot distinguish someone who used occasionally over decades from someone who recently used frequently.

Future studies could improve on that uncertainty by tracking:

• Current frequency, quantity, and THC and CBD content.
• Tobacco use intensity and whether substances are used together or separately.
• Changes in symptoms and cognition before and after changes in use.

These measures could help distinguish exposure effects from differences that existed beforehand. Repeated assessments would also make it easier to investigate whether worsening symptoms precede heavier use, follow it, or move alongside other changes.

Readers interested in how longer-term designs approach these questions can explore MyCannabis coverage of cannabis use, attention, and mental health in young adults. That research involves a different population, but illustrates why tracking changes over time adds information that a single assessment cannot provide.

What the Current Study Leaves Unanswered

The large overall sample is a strength, but the cannabis-only and co-use groups contained just 40 and 48 people. Participants with recent substance abuse or dependence, except tobacco, were excluded, as were those with positive drug screening on the assessment day. The findings therefore do not represent the full range of people with psychosis and problematic cannabis use.

The authors also describe symptom differences as relatively small, with uncertain clinical significance. A statistically detectable association is not automatically a change that meaningfully improves daily life. These limitations make replication in newer, more detailed datasets especially valuable.

A More Precise Path for Cannabis and Mental Health

There is room for optimism in the move toward more specific questions. MyCannabis recently covered experimental research into THC-related brain signaling, which examines biological mechanisms in mice rather than clinical outcomes in people. Such approaches address different parts of the problem and may eventually help explain why responses vary.

For clinical conversations, the new study supports looking at cannabis and tobacco history together, rather than assuming one explains everything. Tobacco cessation remains an important health objective, but these findings do not show that stopping tobacco makes cannabis safe for someone with psychosis.

The defensible promise is better understanding: clearer exposure measurements, more informative comparisons, and care that recognizes individual circumstances. Tobacco is one part of that picture. Accounting for it can move cannabis research beyond competing headlines toward evidence that is more useful to patients, families, and clinicians.

References:

¹ Nelson, A. M., Srinivasan, A., Clementz, B., Gershon, E., Keshavan, M., McDowell, J., Pearlson, G., Tamminga, C., Brady, R. O., Jr., Yassin, W., & Ward, H. B. (2026). Tobacco use moderates associations between cannabis use, psychiatric symptoms, and cognitive performance in psychosis spectrum disorders. Schizophrenia Research, 298, 88–95. https://doi.org/10.1016/j.schres.2026.09.046