Two people walk into a substance use service with the same primary diagnosis. One is a male, the other a female. But how likely is it that the rest of their psychiatric records look the same?
Substance use disorders (SUDs) rarely exist in isolation. Conditions such as depression, anxiety, psychosis and trauma-related disorders commonly occur alongside them, and these comorbidities may influence both the course of an individual’s difficulties and their treatment outcomes. Despite this complexity, integrated care remains limited. Recent estimates suggest that only 7% of US adults with co-occurring substance use and mental health disorder receive integrated treatment (Edmonds et al., 2024).
Previous research suggests that psychiatric comorbidities may not be distributed equally between males and females. For instance, anxiety and trauma-related disorders have been reported more commonly among females, while other disorders show different patterns (Kozak et al., 2021). It remains unclear why these differences occur; however, it could involve a mixture of biological, behavioural and social factors.
Butelman et al. (2026) analysed a large, real-world US clinical dataset to ask whether psychiatric comorbidities differed by sex among people diagnosed with opioid, alcohol or cannabis dependence. By doing this, they aimed to identify patterns that could help us think more carefully about personalised care for people experiencing both substance use and mental health difficulties.
Methods
The study was cross-sectional and used data collected in 2022 from state-funded and state-run mental health services across the US. The sample included adults with a diagnosis of opioid (n=28,808), alcohol (n=23,281), or cannabis dependence (n=5,961).
The researchers examined whether having a secondary diagnosis of anxiety, bipolar disorder, depression, schizophrenia or another psychiatric disorder, or a trauma- and stressor-related disorder differed by sex. A multinomial logistic regression was used to compare the odds of these psychiatric comorbidities, while adjusting for age, race and ethnicity.
This was an analysis of routinely collected clinical data, indicating that it may identify associations between sex and recorded psychiatric disorders; however, it cannot establish what caused these differences.
Results
Across the three substance use groups, a consistent pattern emerged: males had lower adjusted odds of being diagnosed with anxiety, bipolar disorder, depression, and trauma- and stressor-related disorders than females. This pattern remained consistent in the differences in age, race and ethnicity.
Opioid dependence
- For individuals with opioid dependence, anxiety (14.7%) and depression (13.5%) were the most common psychiatric comorbidities.
- Males had lower adjusted odds of anxiety, bipolar disorder, depression and trauma- and stressor-related disorders than females. For instance, Males had 16% lower odds of a trauma- or stressor-related disorder (OR=0.84, 95% CI [0.80 to 0.88]).
- However, the pattern reversed for psychosis: Males had 24% higher odds of schizophrenia or another psychotic disorder (OR=1.24, 95% CI [1.13 to 1.35]).
Alcohol dependence
- Similarly, in individuals with alcohol dependence, depression was the most common comorbidity (21.8%), followed by anxiety (12.0%).
- Males had lower adjusted odds of anxiety, bipolar disorder, depression and trauma- and stressor-related disorders. For instance, the odds of a trauma- or stressor-related disorder were 23% lower among males than females (OR=0.77, 95% CI 0.73 to 0.81).
- Unlike the other two substance groups, there was no significant sex difference in schizophrenia or other psychotic disorders.
Cannabis dependence
- For cannabis dependence, the pattern was slightly different. Depression (18.0%) and schizophrenia or other psychotic disorders (13.4%) were the two most common comorbidities.
- Males had lower adjusted odds of anxiety, bipolar disorder, depression and trauma- and stressor-related disorders with their odds of a trauma- or stressor-related disorder being 32% lower than females’ (OR=0.68, 95% CI [0.61 to 0.75]).
- Males had 12% higher odds of schizophrenia or another psychotic disorder (OR=1.12, 95% CI [1.02 to 1.23]).
Additionally, differences according to race, ethnicity, and age were identified; however, these patterns varied between substance groups and psychiatric diagnoses. Most importantly, sensitivity analyses broadly supported the main sex-based findings.
Overall, one sex did not have “more” psychiatric comorbidity than the other. But rather, the type of recorded psychiatric comorbidity differed by sex, and this pattern was remarkably consistent across opioid, alcohol and cannabis dependence.
Conclusion
The study concluded that there are clear sex-based differences in the types of psychiatric comorbidities recorded among people with opioid, alcohol and cannabis dependence. Across all three substance groups, males had lower adjusted odds of anxiety, bipolar disorder, depression and trauma- and stressor-related disorders than females, while higher odds of psychotic disorders were found among males with opioid or cannabis dependence.
Rather than providing a simple explanation for these differences, the findings highlight the need to better understand the potential interactions between biological, psychological and social mechanisms, with the longer-term goal of improving personalised prevention, diagnosis and treatment.
Strengths and Limitations
A major strength of this study is its size. Data from over 58,000 adults with opioid, alcohol or cannabis dependence were used, which provided more statistical power than many smaller cohort studies.
The use of routinely collected data from state-funded and state-run mental health services across most of the US also offers a valuable picture of psychiatric comorbidity in real-world clinical practice, rather than among the highly selected participants often recruited into research trials.
Furthermore, the analyses adjusted for age, race and ethnicity, rather than treating sex as the only characteristic potentially associated with psychiatric comorbidity. Sensitivity analyses largely supported the main findings, adding some confidence that the observed patterns were reasonably robust.
However, bigger does not necessarily mean better. As a cross-sectional observational study, the findings show associations, not causation. This means that we cannot conclude that sex itself caused the differences in psychiatric comorbidities. The dataset cannot disentangle potential biological differences from social factors such as trauma exposure, stigma, help-seeking behaviour, access to treatment or differences in how symptoms are recognised and diagnosed by clinicians.
There are also important limitations to the dataset. Around 20% of race and ethnicity data were missing and cases with missing values were excluded, while education was missing for 76% of participants and could not be included in the main analyses. The study also considered only one secondary psychiatric diagnosis per person, potentially underrepresenting the complexity of multiple comorbidities commonly seen in clinical practice.
Finally, the sample only included people accessing state-funded or state-run services and may not represent those receiving private care or no treatment at all. The substance-use variable also used DSM-IV “dependence” terminology rather than the current DSM-5 SUD framework. Together, these limitations mean that the findings are valuable for identifying patterns but should not be interpreted as evidence of fixed or universal differences between male and females.
Implications for Practice
Let’s return to the two people who walked into our substance use service. Should their sex change how clinicians assess or treat them? Not necessarily. But this study suggests that it may be one part of a much more complex picture that clinicians should be aware of.
The findings reinforce the importance of looking beyond a person’s primary substance use diagnosis. Psychiatric comorbidities can influence clinical trajectories and treatment outcomes, yet the authors highlight that integrated treatment for co-occurring substance use and mental health disorders remains limited (Edmonds et al., 2024). Comprehensive assessment for co-occurring mental health difficulties therefore remains important when someone presents with a substance use disorder.
However, these findings should not become a shortcut for clinical decision-making. Knowing that females in this dataset had higher adjusted odds of anxiety, depression, bipolar disorder and trauma- and stressor-related diagnoses does not mean clinicians should expect these conditions in every female patient. Likewise, the higher odds of psychotic disorders among males with opioid or cannabis dependence should not lead clinicians to assume that psychosis is inherently a “male” presentation. Instead, these population-level patterns could encourage clinicians to remain alert to the range of possible comorbidities while assessing each person individually.
The study raises concerns about what lies behind the observed differences. Several possibilities, ranging from neurobiological and hormonal mechanisms to differences in trauma exposure, psychological distress, stigma, help-seeking, clinical presentation and access to treatment were identified. However, the dataset cannot tell us which of these explanations, or which combination of them, accounts for the findings.
The associations with race and ethnicity add another layer of complexity. It was suggested that socioeconomic circumstances, exposure to environmental stressors and cultural factors within patient-clinician interactions could contribute to differences in recorded diagnoses. This reinforces the need to understand patients within their broader social and cultural contexts rather than viewing sex as an isolated explanatory factor.
Future research should therefore move beyond asking whether these differences exist and investigate why they exist. Longitudinal research, more detailed assessment of multiple co-occurring diagnoses, and examination of social and clinical factors could help establish what is driving these patterns.
Therefore, it can be concluded that our two hypothetical patients may share the same substance use diagnosis, but their wider mental health needs could look very different. The message for practice is not to make assumptions based on sex, but to make sure that neither person’s individual needs are overlooked.
Statement of interests
Vivien Ciftci declares no conflicts of interest. AI assisted with proofreading and structuring.
Editor
Edited by Éimear Foley. ChatGPT assisted with language refinement and formatting during the editorial phase.
Links
Primary paper
Eduardo Butelman, Yuefeng Huang, Rita Goldstein, Nelly Alia-Klein (2026). Psychiatric comorbidities in substance use disorders: sex-based differences in a national real-world clinical sample. American Journal of Psychiatry, 183(5), 355–363. https://doi.org/10.1176/appi.ajp.20250828
Other references
Edmonds, A., Barham, C., & Brown, J. (2024). Availability and correlates of integrated treatment for people with co-occurring disorders in outpatient behavioral health treatment facilities. Washington, DC: Office of the Assistant Secretary for Planning and Evaluation, US Department of Health and Human Services.
Kozak, K., Smith, PH., Lowe, DJE., Weinberger, AH., Cooper, ZD., Rabin, RA., & George, TP. (2021). A systematic review and meta-analysis of sex differences in cannabis use disorder amongst people with comorbid mental illness. American Journal of Drug and Alcohol Abuse, 47(5), 535–547.
