The technical foundation of the study is the Massachusetts All-Payer Claims Database, or APCD, a statewide repository that aggregates billing records from both public and private insurers. Because claims data capture nearly every reimbursed encounter across the health system, they offer something that traditional psychiatric epidemiology cannot: a view of diagnoses as they are actually recorded in routine care, rather than as they emerge from structured diagnostic interviews in research settings. The researchers analyzed person-year observations spanning 2016 through 2018, focusing on adults aged 21 to 64 who had at least one psychiatric diagnosis and were insured either through Medicaid or through commercial plans. In total, the dataset yielded 2,409,794 person-year observations representing 653,413 unique people with mental health disorders.

Defining multimorbidity in claims data required careful methodological choices. The team identified psychiatric disorders using ICD-10-CM diagnostic codes, the standardized classification system that accompanies every insurance claim, and grouped those codes into broader diagnostic categories. Psychiatric multimorbidity was then defined as the documentation of two or more distinct psychiatric diagnostic categories within a single calendar year. This calendar-year approach matters: it captures the diagnostic landscape a clinician actually faces over a sustained period, rather than a snapshot from a single visit. One important limitation shaped the analysis from the start. Substance use disorder diagnoses were redacted from the database under its privacy policies and therefore had to be excluded, a constraint the authors acknowledge makes their prevalence estimates conservative, since substance use disorders are among the most common co-occurring conditions in psychiatric populations.

The headline numbers are striking in their simplicity. Among the insured adults with a mental health diagnosis, 68.8 percent carried only one psychiatric diagnosis in a given year. Another 17.4 percent carried two, and 13.8 percent carried three or more. Taken together, nearly a third of the population—31.2 percent—met the study’s definition of psychiatric multimorbidity. That figure aligns with a growing body of research suggesting that co-occurrence is the norm rather than the exception in mental illness. Classic epidemiological surveys, including the National Comorbidity Survey Replication, documented high rates of comorbidity decades ago, and more recent work in veterans’ health systems and specialty programs for serious mental illness has repeatedly found that multiple diagnoses cluster together. What distinguishes the new study is its setting: an entire statewide insured population spanning both Medicaid and commercial coverage, rather than a single health system, a specialty clinic, or a volunteer research cohort.

Beyond raw prevalence, the study probed who is most likely to carry multiple psychiatric diagnoses, using multivariable logistic regression to adjust for a range of patient characteristics. The results were consistent and, in places, dramatic. Women had 23 percent higher odds of psychiatric multimorbidity than men, with an adjusted odds ratio of 1.23. People with any medical comorbidity—a physical health condition such as diabetes, heart disease, or chronic lung disease—had 37 percent higher odds than those without, at an adjusted odds ratio of 1.37. Medicaid enrollees had 25 percent higher odds than people with commercial insurance. Each of these associations was estimated with tight confidence intervals, reflecting the enormous sample size, and each points toward a distinct thread of clinical and policy significance.

The most striking association involved serious mental illness. People diagnosed with conditions such as schizophrenia, bipolar disorder, or other severe psychiatric disorders had nearly nine times the odds of psychiatric multimorbidity compared with those without serious mental illness—an adjusted odds ratio of 8.92. This is not merely a statistical curiosity. It means that for the population with the most disabling psychiatric conditions, multiple concurrent diagnoses are close to a defining feature of the illness experience. It also resonates with a 2024 systematic review and meta-analysis in The Lancet Psychiatry, which found that multimorbidity, spanning both psychiatric and physical conditions, is markedly more prevalent among people with severe mental illness than in the general population. The new claims-based findings extend that picture to the diagnostic layer, showing that even within psychiatry itself, severe illness rarely presents in isolation.

The gender pattern echoes a well-documented international literature. Women consistently show higher rates of internalizing conditions such as depression and anxiety, and meta-analytic work has found that women are at greater risk of conditions including obsessive-compulsive disorder, while also being more likely to seek help from general practitioners and to report mental health symptoms. Whether the female excess in documented multimorbidity reflects true differences in disease occurrence, differences in help-seeking and diagnostic capture, or some interaction of both remains an open question in the field. Claims data cannot fully disentangle these forces, since a diagnosis must be both made and billed to appear in the record. Still, the consistency of the association across data sources and countries suggests that the higher multimorbidity burden among women is a real feature of psychiatric epidemiology that treatment systems need to account for.

The insurance-type finding carries its own weight. Medicaid enrollees, who tend to have lower incomes and higher rates of disability, showed consistently higher odds of psychiatric multimorbidity than commercially insured adults even after statistical adjustment. This aligns with prior research showing that insurance status shapes both access to mental health care and the intensity with which conditions are diagnosed and treated, and with studies documenting that Medicaid is the dominant payer for people with serious mental illness. The association may partly reflect greater clinical complexity among Medicaid populations, greater contact with the health system and therefore more opportunities for additional diagnoses to be recorded, or both. Either way, the pattern underscores that the burden of complex psychiatric presentations falls disproportionately on the publicly insured, with direct implications for how state Medicaid programs budget, design networks, and structure integrated care.

Why does all of this matter clinically? Psychiatric multimorbidity is increasingly understood as a marker of clinical complexity, associated with greater symptom burden, poorer quality of life, higher service utilization, and worse outcomes. Research in patients with schizophrenia has linked mental health multimorbidity to substantially poorer quality of life, and studies of veterans with post-traumatic stress disorder have documented extensive diagnostic overlap. Yet most mental health care remains organized around single-disorder treatment protocols, with separate pathways for depression, anxiety, PTSD, and psychosis. The authors argue that their findings point toward integrated and patient-centered treatment approaches—models that treat the person and their full diagnostic profile rather than marching through diagnoses one at a time. There is also a growing transdiagnostic movement in psychiatry, which seeks interventions that cut across diagnostic boundaries precisely because co-occurrence is so common.

The study also carries a methodological lesson for the field. Structured diagnostic interviews, the gold standard of psychiatric epidemiology, capture small, carefully selected samples. Claims data capture everyone who touches the health system, at the cost of depending on the accuracy and completeness of billing diagnoses. By demonstrating that large-scale administrative data can be used to characterize psychiatric multimorbidity across an entire state, the researchers have opened a path for surveillance that could be repeated over time and replicated in other states with all-payer databases. Because substance use disorders were excluded, the true prevalence of multimorbidity is almost certainly higher than reported here. Even so, the message is difficult to escape: in the real world of American health care, mental illness most often arrives with company, and health systems designed for one diagnosis at a time are increasingly out of step with the patients they serve.

Subject of Research: Prevalence and correlates of psychiatric multimorbidity in a statewide insured population

Article Title: Measuring and Characterizing Psychiatric Multimorbidity Among Individuals with Mental Illness

Article References: Measuring and Characterizing Psychiatric Multimorbidity Among Individuals with Mental Illness. (n.d.). https://doi.org/10.1007/s10597-026-01675-w

Image Credits: AI Generated

DOI: 10.1007/s10597-026-01675-w

Keywords: psychiatric multimorbidity, mental illness, comorbidity, all-payer claims database, Medicaid, serious mental illness, ICD-10-CM, epidemiology, health services research, integrated care, substance use disorder, Massachusetts