Donnelly · Psychoneuroendocrinology 2022 · prospective birth cohort study · n=3258

Childhood immuno-metabolic markers and risk of depression and psychosis in adulthood: A prospective birth cohort study.

Cited 10 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective longitudinal birth cohort study

PubMed 35286909 · doi:10.1016/j.psyneuen.2022.105707 · record verified 2026-08-27

What was done

This prospective birth cohort study assessed 3,258 participants to evaluate whether metabolic hormones (leptin, adiponectin, insulin) at age 9 predicted depression- and psychosis-spectrum outcomes at age 24. Researchers also evaluated nine immuno-metabolic biomarkers (leptin, adiponectin, insulin, interleukin-6, C-reactive protein, LDL, HDL, triglycerides, and BMI) structured into an exploratory bifactor model comprising a general immuno-metabolic factor and three specific subfactors (adiposity, inflammation, and insulin resistance).

What was found

Childhood leptin was significantly associated with adult depressive episodes (adjusted odds ratio [aOR] = 1.31; 95% CI, 1.02–1.71) and negative symptoms (aOR = 1.15; 95% CI, 1.07–1.24), but not positive psychotic symptoms. The general immuno-metabolic factor was associated with atypical depressive symptoms (aOR = 1.07; 95% CI, 1.01–1.14) and psychotic experiences (aOR = 1.21; 95% CI, 1.02–1.44). The adiposity factor was associated with negative symptoms (aOR = 1.07; 95% CI, 1.02–1.12). Point estimates were generally higher in women (though 95% credible intervals overlapped with men); in women, the inflammatory factor specifically associated with depressive episodes (aOR = 1.27; 95% CI, 1.03–1.57).

Why it matters

These findings suggest that peripheral immuno-metabolic alterations present in mid-childhood precede adult psychiatric symptoms by 15 years, supporting a potential developmental and biological role for metabolic-inflammatory pathways in mood and psychotic disorders.

Limits

The observational design precludes causal inference, and residual confounding across the 15-year follow-up cannot be ruled out. Observed effect sizes were relatively modest (aORs mostly ranging between 1.07 and 1.31). The bifactor model was exploratory, and the abstract does not report cohort attrition rates or intermediate longitudinal biomarker trajectories between ages 9 and 24.

Cited by