Abstract:Understanding the biological mechanisms behind multimorbidity patterns in adolescence is important as they may act as intermediary risk factor for long-term health. We aimed to explore relationship between prenatal exposures and adolescent's psycho-cardiometabolic intermediary traits mediated through epigenetic biomarkers, using structural equation modeling (SEM). We used data from mother–child dyads from pregnancy and adolescents at 16–17 years from two prospective cohorts: Northern Finland Birth Cohort 1986 (NFBC1986) and Raine Study from Australia. Factor analysis was applied to generate two different latent factor structures: (a) prenatal exposures and (b) adolescence psycho-cardiometabolic intermediary traits. Furthermore, three types of epigenetic biomarkers were included: (1) DNA methylation score for maternal smoking during pregnancy (DNAmMSS), (2) DNAm age estimate PhenoAge and (3) DNAm estimate for telomere length (DNAmTL). Similar factor structure was observed between both cohorts yielding three prenatal factors, namely BMI (Body Mass Index), SOP (Socio-Obstetric-Profile), and Lifestyle, and four adolescent factors: Anthropometric, Insulin-Triglycerides, Blood Pressure, and Mental health. In the SEM pathways, stronger direct effects of F1 prenatal -BMI (NFBC1986 = β : 0.27; Raine = β : 0.39) and F2 prenatal -SOP ( β : −0.11) factors were observed on adolescent psycho-cardiometabolic multimorbidity. We observed an indirect effect of prenatal latent factors through epigenetic markers on a psycho-cardiometabolic multimorbidity factor in Raine study ( P < 0.05). The present study exemplifies an evidence-based approach in two different birth cohorts to demonstrate similar composite structure of prenatal exposures and psycho-cardiometabolic traits (despite cultural, social, and genetic differences) and a common plausible pathway between them through underlying epigenetic markers.

Developmental origins of psycho-cardiometabolic multimorbidity in adolescence and their underlying pathways through methylation markers: a two-cohort study