One‐year incidence, predictors, and accuracy of prediction of suicidal thoughts and behaviors from the first to second year of university

Abstract

Background Research is scarce on the prospective predictors of first onset suicidal thoughts and behaviors (STB) and the accuracy of these predictors in university students, particularly in low‐and‐middle income countries. Therefore, we assessed the 1‐year incidence of STB among first‐year students, a broad range of prospective predictors of STB incidence, and evaluated the prediction accuracy of a baseline multivariate risk prediction model to identify students at highest risk for STB onset over the subsequent 12 months. Methods Students (n = 3238) from 13 universities in Mexico completed an online survey developed for the World Mental Health International College Student Surveys in their first year and again 12 months after. We ran generalized linear models and receiver operator curves. Results The 1‐year incidence of suicidal ideation, plan and attempt was 8.53%, 3.75%, and 1.16%, respectively. Predictors in final models were female sex (ideation only), minority sexual orientation (ideation only), depression, eating disorders, ADHD (ideation and plan), ongoing arguments or breakup with a romantic partner (ideation only), emotional abuse (ideation only), parental death (ideation, plan), not Catholic/Christian (ideation, plan), not having someone to rely on, psychotic experiences (plan only), and insufficient sleep (attempt only). Prediction accuracy for ideation, plan and attempt was area under the curve = 0.76, 0.81 and 0.78, respectively. Targeting the top 10% of students at highest risk could reduce STB in the subsequent year up to 36%. Conclusions By assessing these risk/protective factors in incoming students we identified students at greatest risk for developing STB to whom suicide prevention strategies could be targeted.

Frequency and machine learning predictors of depressive, anxiety, obsessive-compulsive symptoms, and suicidal ideation among university students

“…Noteworthily, suicidal ideation at baseline could be leveraged by the machine to predict suicidal ideation after six months, but it did not yield a strictly statistically significant role. To our knowledge, there have been several studies implementing machine learning models leveraging previous suicidal ideation to predict future suicidal ideation (Benjet et al, 2022;Liao et al, 2022;Malone et al, 2021), although timeframes from baseline to follow-up did not coincide with our six-month timeframe. Substantially, previous suicidal ideation is an important is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint…”

“…Noteworthily, suicidal ideation at baseline could be leveraged by the machine to predict suicidal ideation after six months, but it did not yield a strictly statistically significant role. To our knowledge, there have been several studies implementing machine learning models leveraging previous suicidal ideation to predict future suicidal ideation (Benjet et al, 2022; Liao et al, 2022; Malone et al, 2021), although timeframes from baseline to follow-up did not coincide with our six-month timeframe. Substantially, previous suicidal ideation is an important factor, posing a greater risk of future suicidal ideation and suicidal behavior (Bafna et al, 2022), but in our analysis did not reach canonical statistical significance (although it should be taken into account that the machine learning model would lose a great proportion of its accuracy without data from previous suicidal ideation – Figure 3A).…”

Frequency and machine learning predictors of severe depressive symptoms and suicidal ideation among university students

“…To our knowledge, there have been several studies implementing machine learning models leveraging previous suicidal ideation to predict future suicidal ideation (Benjet et al. , 2022 ; Liao et al. , 2022 ; Malone et al.…”

Developing a risk prediction model for death at first suicide attempt—Identifying risk factors from Thailand’s national suicide surveillance system data

“…A few studies on first-onset suicidal thoughts and behaviors have been conducted among college students as this group appears to be at the greatest risk of developing first-onset suicidal thoughts and behaviors. Two studies among college students found that adverse experiences such as dating violence and physical abuse before the age of 17, betrayal by someone other than the partner [ 19 ], ongoing arguments or breakup with a romantic partner, emotional abuse, parental death and lack of a reliable support system [ 20 ] were strong predictors for first-onset suicidal thoughts and behaviors.…”