Immigration · Primary-source case analysis

Uribe v. Garland: A Mental-Health Social Group Needed Record-Defined Boundaries

By Aneeta Mishra and Timothy Chan

Published August 11, 2026 · Updated October 20, 2026

Decision: Uribe Andrade v. Garland, No. 21-1244 (9th Cir. Mar. 1, 2024).

The Ninth Circuit denied Miguel Angel Uribe Andrade’s petition because the record did not give his proposed mental-health social group sufficiently clear boundaries, and the evidence did not compel CAT protection. The decisive gap was not the absence of any diagnosis. It was the missing definition of a behavioral modifier used to identify the group, followed by separate evidentiary gaps in his predicted path to torture. Uribe Andrade v. Garland, No. 21-1244 (9th Cir. Mar. 1, 2024), opinion, pages 4, 11–22.

Criminal removability did not eliminate review of legal questions

Uribe conceded removability based on controlled-substance convictions. Section 1252(a)(2)(C) restricted review of that removal order, but section 1252(a)(2)(D) preserved constitutional claims and questions of law. He disclaimed challenges to the factual findings supporting the social-group ruling and instead argued that the agency had applied the law incorrectly. The court therefore reviewed that legal question de novo. Opinion, pages 9–10.

CAT review was different. Applying Nasrallah, the court could review the factual findings underlying the CAT denial despite the criminal-removal restriction. Substantial-evidence review still required Uribe to show that the record compelled a contrary result; it did not permit the court to choose its preferred interpretation of conflicting evidence. Opinion, pages 18–19.

Particularity required a clear, record-supported boundary

The proposed group was “Mexicans with mental health disorders characterized by psychotic features who exhibit erratic behavior.” A mental-health evaluator diagnosed major depression with psychotic features and described the importance of treatment. But neither the record nor his providers defined the additional phrase “erratic behavior” or used it to describe his condition or symptoms. Uribe did not challenge those factual findings on appeal. Opinion, pages 6, 8–9, 11–12.

That distinction drove the result. Particularity asks whether a group has discrete, definable boundaries—a workable way to determine membership. Conduct labeled erratic could vary in frequency, duration and character. Adding the phrase might narrow a group without establishing a clear boundary. The Board was not required to invent a lay definition to fill the missing evidence. It assumed social distinction for its analysis and rejected the group on particularity. Opinion, pages 8–9, 11–13, 17–18.

The court distinguished a case with actual clinical evidence

Uribe relied on Acevedo Granados, which also involved a proposed group including erratic behavior. The panel explained what that earlier decision actually addressed: The agency had disregarded a professionally documented intellectual-disability diagnosis and substituted its own lay assessment. It remanded for further factfinding; it did not decide that every similarly worded group was legally cognizable or resolve the meaning of the behavioral phrase. Here, by contrast, the agency’s problem was the absence of record evidence defining the disputed modifier. Opinion, pages 13–16.

The panel also answered Uribe’s reliance on the Fourth Circuit’s Temu decision. That case involved bipolar disorder and expert evidence linking visibly erratic behavior to a recognized perception of demonic possession in Tanzania. Uribe proposed a broader mental-health formulation and had no equivalent country-specific evidence defining the disputed term. The Board considered his group as formulated, rather than improperly demanding that the behavioral modifier qualify as a group by itself. Opinion, pages 16–18.

The CAT theories required more than generalized risk

The institutionalization theory depended on a chain: Inability to obtain needed treatment in Mexico, deterioration leading to involuntary commitment, and torture in an institution. The court required the necessary links to be more likely than not. The record did not compel a finding that medication or treatment would be unavailable. The affordability argument rested on a conclusory assertion about employment and family support, and Uribe had never previously been institutionalized. The agency could therefore find the predicted commitment insufficiently established. Opinion, pages 19–20.

There was a separate intent problem even assuming commitment. Reports documented poor institutional conditions, but the record also supported the agency’s inference that limited resources, rather than a specific intent to inflict severe suffering, explained them. Under deferential review, that alternative supported denial on this record. Evidence describing conditions and evidence establishing the legally required intent did different work. Opinion, pages 20–21.

The tattoo and former-gang-affiliation theory also lacked a sufficiently individualized connection. General evidence of cartel violence did not compel a finding that Uribe faced a particularized threat, or that officials would consent or acquiesce. The panel then addressed his actual aggregate-risk argument: The immigration judge expressly considered the risks together, and the Board adopted and cited that discussion. The court found no failure to consider their combined probability and denied the petition. Opinion, pages 21–22.

Key takeaways

• Compare the exact proposed group with the clinical report. A diagnosis may establish a condition without defining an additional behavioral boundary; identify what evidence supplies each component.
• Preserve the group and its supporting evidence in the agency record. Our Carmona-Sanchez analysis addresses the separate problem of introducing a new group or materials for the first time in court.
• For a treatment-based CAT theory, identify evidence of actual treatment access, affordability, likely institutionalization and the source of the feared intentional harm. General descriptions of poor facilities do not answer every link.
• Address all claimed sources of torture together while documenting the individual threat and official involvement or acquiescence for each theory.
• Separate the legal social-group challenge from CAT factual review when criminal-removal restrictions apply. This published March 2024 opinion denied review on the record presented.