Laika Aguinaldo
University of California-San Diego, CA·Psychology·Early-career
Suicide is the second leading cause of death among young people aged 10 to 19, and rates are rising. Better early detection is urgently needed.
- Fields
Psychology
Public Health & Epidemiology
Mathematics & Statistics
- Topics
youth suicide
suicide risk prediction
suicidal thoughts and behaviors
- Grants
- 1 known grant, through Jul 2031
What they work on
- Assess and correct measurement non-invariance, where measures behave differently across groups, in brain, psychological, health record and social predictors of youth suicide risk, using Moderated Nonlinear Factor Analysis on ABCD Study data.
- Predict future suicidal thoughts and behaviors from ages 9 to 18 using interpretable machine learning methods such as random forest, elastic net, support vector machines and extreme gradient boosting, with nested and forward-chaining cross-validation.
- Improve model calibration and accuracy within subgroups using methods such as Trans-Balance, Platt scaling and stratified modeling, and validate the models externally in the National Consortium on Alcohol and Neurodevelopment in Adolescence-Adulthood cohort.
More about the research
Machine learning models can predict suicidal thoughts and behaviors with reasonable accuracy across a whole population. They often fail, however, for particular groups defined by sex, income, developmental stage, region or race and ethnicity. One reason is measurement non-invariance, meaning a questionnaire or other measure does not capture the same thing equally well in every group. This can cause youth to be misclassified and miss chances for help. Preliminary findings from the Adolescent Brain Cognitive Development (ABCD) Study also show that brain-based measures alone predict risk poorly.
Why it matters
Many youth at risk of suicide are not identified in time, and current prediction models work less well for some groups. A risk model that is accurate and well-calibrated across sex, income, development stage, region and race/ethnicity could support suicide risk assessment in research and clinical settings. It also lays groundwork for future use with real-world electronic health records, via the EPIC Cosmos network covering over 180 million patients.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Jocelyn Meza
University of California-Los Angeles, CA·Psychology
Chart the longitudinal trajectories of suicidal thoughts and behaviors from age 10 to 28 in 674 Mexican-origin youth from the California Families Project.
Carolina Makowski
University of California-San Diego, CA·Neuroscience·Early-career
Identify behavioral and neuroimaging correlates of eating disorders in youth from the ABCD Study (ages 11–14) and in a clinical sample of adolescent girls in treatment (ages 13–18), using measures of brain shape and whole-brain microstructure.
Andrea Baldelli
University of North Carolina at Chapel Hill, NC·Psychology·Early-career
Refine questionnaires, interviews and behavioral tasks that measure social motivation and suicidal thoughts and behaviors so they are valid for a diverse group of preteen girls aged 8 to 11.
The school
University of California-San Diego: 84 labs on PI Pages, $1.9B a year on research, 637 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.