A Retrospective Study on Mucous Stool Excretion in Giant Pandas: Statistical Analysis Based on CCRCGP Data

This retrospective multicenter study integrated individual- and event-level records from 190 captive giant pandas managed at four study sites of the China Conservation and Research Center for the Giant Panda (CCRCGP) during 2024 to characterize patterns of mucous stool excretion and associated phenotypic variation. Logistic and negative binomial regression models were used to evaluate individual-level associations with event occurrence and annual event frequency, while generalized estimating equations (GEEs) were applied to monthly and event-level analyses to account for repeated observations within pandas. Overall, 132 of 190 pandas (69.5%, 95% CI, 62.6-75.6%) experienced at least one mucous stool event, with 674 events recorded during the study year. Monthly event frequency differed significantly overall (Poisson GEE, p = 0.0015), with the largest observed event count occurring in July, although no individual month-versus-January comparison remained significant after Holm correction. The monthly probability of an individual panda experiencing at least one event did not differ significantly (p = 0.295). Mucous stool weight showed no significant monthly variation (p = 0.453) but differed markedly by physical phenotype, compared with jelly-like stools, lard-like and mixed jelly/lard stools were 1.318- and 2.133-fold heavier, respectively. An exploratory inverse association between body weight and event occurrence persisted after adjustment for age and sex (adjusted OR per kg = 0.961, 95% CI, 0.930-0.992, p = 0.015). Recurrently affected individuals contributed disproportionately to the annual event burden, but did not exclusively account for the elevated July count. These findings indicate that within-year variation primarily reflected changes in event burden rather than the proportion of affected pandas, while visible phenotype was strongly associated with the amount of mucous material excreted. The disproportionate contribution of recurrently affected individuals highlights the value of targeted longitudinal monitoring in veterinary and husbandry management, while multi-year prospective surveillance is needed to confirm reproducible temporal patterns and clarify underlying mechanisms.