Measuring levels of fatty acids, phospholipids, and sterols in erythrocyte membranes may help predict which ultra-high-risk (UHR) individuals will experience conversion to full psychosis, new research suggests.
Although analyzing individual lipids, such as linoleum acid, or phospholipids was somewhat predictive of conversion, it was only when all three groups were examined together that the researchers were able to significantly predict psychosis conversion.
“We believe we can use membrane lipids for personalized medicine in psychiatry, especially in psychosis,” said lead researcher Ariel Frajerman, MD, Centre Hospitalier Sainte Anne, Paris, France.
The findings were presented at the virtual Congress of the Schizophrenia International Research Society (SIRS) 2021.
Novel Avenue of Research
Frajerman noted that UHR individuals already have some symptoms of psychosis and that 20% to 40% of them will experience conversion to full psychosis in the next 3 years.
“The question is, which ones?” he said.
Membrane lipids are an interesting avenue of research compared with plasma lipids because erythrocytes have a “life expectancy of 120 days,” so measurements are considered to be more stable over time, Frajerman said.
Membrane lipids include the polyunsaturated fatty acids omega-3, omega-6, and omega-9; the phospholipids phosphatidyl choline, phosphatidyl ethanolamine, phosphatidyl serine, and sphingomyeline; and the sterols cholesterol and cholestanol.
“They have a very important role for the biochemical and physical properties of the membrane,” Frajerman said. He noted that their presence is “good…not like in the plasma.”
Previous attempts to use lipid levels in erythrocyte membranes to predict psychosis conversion yielded conflicting results because of the low number of patients who experienced conversersion or because the studies focused on specific lipids, he noted.
The researchers examined erythrocytes from a cohort of 61 UHR individuals in France in the ICAAR study. Among the participants, 29 experienced conversions, and 32 did not.
For erythrocyte assessment, the investigators combined liquid chromatography with mass spectrometry, “which is the gold standard” for lipid analysis, Frajerman said.
Combining Lipids Key
Results showed a high correlation between levels of fatty acids, phospholipids, and sterols.
Regarding only fatty acids, the area under the receiving operating characteristic curve (AUC) for predicting psychosis conversion was 0.653, which is “not very good,” Frajerman noted.
Linoleic acid alone was a better predictor of psychosis conversion, at an AUC of 0.654. The ratio of cholestanol and cholesterol was not as good, at an AUC of 0.607.
Phospholipids had a similar ability to predict psychosis conversion, at an AUC of 0.645.
Combining linoleic acid and phospholipids improved the ability to predict psychosis conversion, at an AUC of 0.697. Adding the cholestanol/cholesterol ratio improved the predictive power further, at an AUC of 0.728 (P = .004).
Frajerman noted that their study included 29 patients who experienced conversion, which was “not a lot,” but that the cohort is all from the same country. This is “important because it has already been demonstrated that diet over the long term has an impact on fatty acids, [so there] will be differences between countries,” he said.
Study limitations include the fact that there was no control group, there were no dietary data, and there was no replication cohort. Frajerman said future research will include food questionnaires.
Following the presentation, session co-chair Alexis Cullen, PhD, research fellow and honorary lecturer at the Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, United Kingdom, asked whether there were any correlations between membrane lipids and symptoms.
Frajerman said there was an association between membrane lipid levels at baseline and reduction in Social and Occupational Functioning Assessment Scale scores 1 year later.
Commenting on the study for Medscape Medical News, David R. Cotter, MD, PhD, Royal College of Surgeons in Ireland and the Education and Research Center, Beaumont Hospital, Dublin, Ireland, said the findings were “interesting” but need to be replicated
If replicated, clinicians would then “need to consider adding these lipid predictors to protein predictors,” such as the proteomic biomarkers identified in a recent study published in JAMA Psychiatry, said Cotter, who was not involved with the research.
He noted that two key questions remain: What is the biological basis of the membrane lipid signature identified in the study, and does it reflect a chronic inflammatory state?
Also commenting, Scott R. Clark, MD, PhD, head of discipline of psychiatry, University of Adelaide, Adelaide, Australia, told Medscape Medical News that previous studies have explored associations between lipids and psychosis.
These include an analysis of plasma lipids in a large number of children and adolescents from the Avon Longitudinal Study of Parents and Children and a study of high-risk individuals whose cases were followed for 5 years. That study used clustering and machine learning to identify lipid profiles linked to psychosis onset.
Clark believes there is potential utility of membrane lipids to personalize care by concentrating on UHR individuals most likely to experience transition to psychosis. “But I think the relationships are more complex and that multi-lipid signatures are likely to be more useful,” he said.
There are “numerous models for psychosis prediction” that combine a range of modes of biological and clinical data, because “all biological predictors have small effect size and need to be combined together to achieve reasonable accuracy,” said Clark.
“Much larger sample sizes are needed to confirm these relationships,” he said. He notied that the National Institute of Mental Health has launched a number of such studies as part of the Accelerating Medicines Partnership. The goal is to identify individuals at increased risk and develop targeted interventions.
The study authors and commentators have reported no relevant financial relationships.
Congress of the Schizophrenia International Research Society (SIRS) 2021: Abstract 3007124. Presented April 19, 2021.