Apolipoprotein E genotype, inflammatory biomarkers, and non-psychiatric multimorbidity contribute to the suicidal ideation phenotype in older age. The Salus in Apulia Study
Madia Lozupone,Rossella Donghia,Rodolfo Sardone,Anita Mollica,Giuseppe Berardino,Luisa Lampignano,Chiara Griseta,Roberta Zupo,Fabio Castellana,Ilaria Bortone,Vittorio Dibello,Emanuela Resta,Roberta Stallone,Davide Seripa,Antonio Daniele,Vincenzo Solfrizzi,Mario Altamura,Antonello Bellomo,Francesco Panza
DOI: https://doi.org/10.1016/j.jad.2022.09.046
2022-12-15
Abstract:Background: Possible relationships between suicidal ideation and biopsychosocial predictors in older age are unclear. In the population-based Salus in Apulia Study, we investigated the relationships among biomarkers, socio-demographic, psychopathological, inflammatory and metabolic characteristics and suicidal ideation in 1252 older subjects. Methods: Suicidal ideation was evaluated with the brief version of the Columbia-Suicide Severity Rating Scale. Apolipoprotein E (APOE) genotype and inflammatory profile [interleukin (IL)-6, tumor necrosis factor (TNF)-α, C-reactive protein (CRP)] were evaluated. A machine learning algorithm, the Random Forest (RF), selected potential biopsychosocial factors associated to suicidal ideation. Results: Suicidal ideators accounted for 2.32 % of subjects, were female, smokers, and obese with multimorbidity. After adjusting for age, gender, education and social dysfunction, logistic regression analyses revealed that suicidal ideation was associated to late-life depression (LLD) (odds ratio:21.71,95 % confidence interval:9.22-51.14). In the full RF model, asthma was the most important contributor to suicidal ideation. In the final RF model, education, age, and mild cognitive impairment followed by gender and global cognition were considered the most important contributors. Among biomarkers, in the final RF model, IL-6 followed by TNF-α, APOE ε4 allele presence, CRP and high-density lipoprotein cholesterol contributed most to suicidal ideation. Limitations: A relatively small number of older subjects with suicidal ideation (2.3 %); we did not distinguish between active and passive suicidal ideation. Conclusions: Although LLD is a strong determinant of suicidal ideation, other non-psychiatric factors, i.e., serum inflammation biomarkers, APOE ε4 allele, and multimorbidity, should be taken into account when evaluating a suicidal ideation phenotype in older age.