Background: The Ontario Marginalization Index (ON MARG) is widely used to assess health inequalities in Ontario by measuring four dimensions of marginalization at the dissemination area (DA) level. However, averaging these dimensions into an overall deprivation score can obscure important information, in particular information on intersectionality of material and social deprivation and race and immigrant status. Objective: To use machine learning algorithms to uncover relationships among the four ON-MARG dimensions across DAs as demographic clusters and to compare the use of these clusters to understand and map marginalization and to describe health inequities. Methods: We applied K-means clustering to 2021 ON-MARG data on the four On-MARG dimensions Households and Dwellings (HD), Material Resources (MR), Age and Labour Force (AL), and Racialized and Newcomer Populations (RN) across 20,123 DAs. We then compared these clusters to ON-MARG average index scores in terms of mapping marginalization in Toronto and examined how these clusters were associated with inequities in mental health service as compared specific dimensions of the ON-MARG index. Results: We identified four clusters: (1) Advantaged White Canadians, (2) Disadvantaged White Canadians, (3) Advantaged Visible Minorities and Immigrants, and (4) Disadvantaged Visible Minorities and Immigrants. The clustering approach revealed nuanced patterns not captured by the ON-MARG summary scores alone. Disadvantaged White Canadians exhibited the highest outpatient mental health visit rates, particularly among females (250 to 300 visits per 100,000). Disadvantaged Visible Minorities and Immigrants followed with elevated rates, while both advantaged clusters showed significantly lower utilization. The clusters provided better discrimination of health service disparities than ON-MARG quintiles alone, highlighting that disadvantaged groups, regardless of racial composition, had higher rates of mental health service use. Conclusions: Combining ON-MARG with machine learning clustering offers a more comprehensive understanding of marginalization's intersectionality, revealing disparities in health service utilization not apparent from the index alone. This approach underscores the need for targeted, intersectional policies to address the specific needs of diverse populations, ultimately contributing to more equitable healthcare interventions in Ontario.

The authors have declared no competing interest.

Institute for Pandemics, University of Toronto

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

University of Toronto IRB

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes

CENSUS data and ON-Marg data are publicly available, health care data is available trough the Ontario Health Data Platform.