The salivary proteome reflects some traits of dietary habits in diabetic and non-diabetic older adults

• Published in: European journal of nutrition 2021-12, Vol.60 (8), p.4331-4344
• Main Authors: Chambon, Christophe, Neyraud, Eric, Sayd, Thierry, Bros, Pauline, Di Biagio, Romane, Hyvrier, Frank, Féart, Catherine, André, Perrine, Rodriguez-Artalejo, Fernando, Lopez-Garcia, Esther, Garcia-Esquinas, Esther, Gomez-Cabrero, David, Proctor, Gordon, Morzel, Martine
• Format: Article
• Language: English
• Subjects: Chemistry; Chemistry and Materials Science; Degranulation; Diabetes; Diabetes mellitus (non-insulin dependent); Diet; Epidemiology; Food and Nutrition; Food intake; Food selection; Gene set enrichment analysis; Life Sciences; Liquid chromatography; Mass spectroscopy; Meat; Nutrition; Older people; Original Contribution; Prediction models; Protein composition; Proteomes; Saliva
• ISSN: 1436-6207; 1436-6215
• DOI: 10.1007/s00394-021-02584-2

Purpose Objective markers of usual diet are of interest as alternative or validating tools in nutritional epidemiology research. The main purpose of the work was to assess whether saliva protein composition can reflect dietary habits in older adults, and how type 2 diabetes impacted on the saliva-diet correlates. Methods 214 participants were selected from 2 European cohorts of community-dwelling older adults (3C-Bordeaux and Seniors-ENRICA-2), using a case–control design nested in each cohort. Cases were individuals with type 2 diabetes. Dietary information was obtained using the Mediterranean Diet Adherence Screener (MEDAS). Saliva was successfully obtained from 211 subjects, and its proteome analyzed by liquid chromatography–tandem mass spectrometry. Results The relative abundance of 246 saliva proteins was obtained across all participants. The salivary proteome differed depending on the intake level of some food groups (especially vegetables, fruits, sweet snacks and red meat), in a diabetic status- and cohort-specific manner. Gene Set Enrichment Analysis suggested that some biological processes were consistently affected by diet across cohorts, for example enhanced platelet degranulation in high consumers of sweet snacks. Minimal models were then fitted to predict dietary variables by sociodemographic, clinical and salivary proteome variables. For the food group «sweet snacks», selected salivary proteins contributed to the predictive model and improved its performance in the Seniors-ENRICA-2 cohort and when both cohorts were combined. Conclusion Saliva proteome composition of elderly individuals can reflect some aspects of dietary patterns.