Predictors of Urinary Neopterin Concentration in Primates: A Multilevel Meta-Analysis

Abstract

Ecoimmunological research in primates increasingly uses urinary neopterin as a noninvasive biomarker of cell-mediated immune activation, but primary studies often disagree on the magnitude and sometimes direction of neopterin-predictor associations, and modest sample sizes limit single-study inferences. We synthesized this literature through a systematic review and multilevel meta-analysis, extracting 83 effect sizes from 13 studies encompassing eight species and 12 populations. Effects were modeled as partial correlations using a three-level random-effects structure with effect sizes nested within studies and species as a crossed random effect, and assigned to six predictor categories (environment, individual characteristics, health, behavior, storage, and individual-by-environment), four individual predictor labels (age, sex, seasonality, infection), and two study designs (naturalistic vs. structured). The overall pooled effect was small but statistically significant. All heterogeneity resided at the effect-size level; the between-study and between-species variance components were estimated at zero. Predictor category moderated the association: environmental predictors yielded the largest effect, followed by individual characteristics, while health, behavior, and storage effects did not differ from zero. The age-specific estimate was consistent in direction across 10 studies, supporting urinary neopterin's value for studying immunosenescence. Naturalistic and structured designs returned similar estimates. A collapsed-population sensitivity analysis confirmed that no single population drove the results, and no evidence of publication bias was detected. Ecological context is the primary aggregate-level predictor of urinary neopterin variation, and neopterin-predictor relationships are consistent across primate species, supporting its utility as a standardized comparative biomarker for primate ecoimmunology.