Developmental intestinal microbiome alterations in canine fading puppy syndrome: a prospective observational study

Fading puppy syndrome (FPS) is a lethal, poorly understood condition in neonatal dogs, responsible for substantial early-life mortality, often within the first week post-partum. Proposed etiologies include dystocia, infections, failure of passive transfer, parasitism, and husbandry factors, with bacterial infections and septicemia considered major contributors. Given evidence from human neonatology that early-life gut microbiome disruption associates with diseases such as necrotizing enterocolitis and late-onset sepsis, the authors hypothesized that early intestinal microbiome composition in puppies is associated with and potentially predictive of FPS. The study was designed to quantify FPS occurrence, identify associated clinical and environmental risk factors, characterize the early rectal microbiome, compare microbiomes of healthy versus fading puppies, and evaluate whether early microbiome signatures predict FPS.

The paper reviews that canine neonatal mortality can affect up to 30% of litters, with FPS presenting non-specific signs and near-100% mortality. Bacterial infections, often maternal-to-neonatal transmission or via intestinal translocation, are implicated as major causes of death. In humans, perturbations in neonatal gut microbiome development are linked to conditions like necrotizing enterocolitis and sepsis. Early gut colonization is influenced by delivery mode, maternal health, milk, antibiotic exposure, and gestational age. In dogs, sequencing-based studies of neonatal gut microbiome are scarce, though prior work shows temporal instability, inter-individual variability, and increasing richness over early life. No prior studies had assessed links between maternal/neonatal microbiomes and canine neonatal disease, motivating this investigation into FPS.

Design: Prospective observational study (2018–2019), institutional ethical approval KSVM-VTH/18_2017, owner consent obtained. Cohort: Pure-bred dams (mostly vaginal deliveries; 1 Cesarean) and their litters from kennels/breeders. Exclusions: dams on antibiotics within 12 months prepartum; stillborn puppies; puppies dying within 24 h PP; puppies given milk replacer or antibiotics before swabbing; congenital defects or non-FPS conditions. Clinical data: Recorded dam demographics, parity, breeding history, parturition data, husbandry, season. Puppies underwent physical exam (vital signs, weight, sex) and activity scored (0 low, 1 moderate, 2 normal). Outcomes tracked; FPS defined as progressive deterioration within 2 weeks PP with weakness, decreased nursing, hypothermia, dehydration, hypoglycemia, or other clinical signs. Sampling: Rectal swabs collected on Day 1 and Day 8 PP using sterile swabs; transported on dry ice within 2 h; stored at −80°C. Rectal temperature measured; weights recorded. Necropsy performed when feasible within 2 h of death. Therapy protocol for FPS: Supportive care including correction of hypoglycemia, tube feeding (Royal Canin Babydog Milk), hydration (IV fluids as needed), warming for hypothermia, oxygen for hypoxia, and clinician-directed antibiotics (amoxicillin–clavulanate 15 mg/kg q12h IV/PO) and probiotics. Microbiome lab methods: DNA extraction with PureLink microbiome kit with bead-beating. 16S rRNA gene V4 region PCR (515F-806R barcoded primers; PrimeSTAR Max). Amplicon cleanup (AMPure XP), quantification (PicoGreen), equimolar pooling, Illumina MiSeq sequencing. Bioinformatics: QIIME2 v2019.4 pipeline. Features grouped at ≥99% similarity; taxonomy assigned with GreenGenes. Chimera removal via DADA2 (trunc length 185). Filtered samples with <500 features and features with total frequency <4. Rarefaction to 2649 sequences/sample. Alpha diversity (Shannon) and beta diversity (Jaccard; UniFrac consistent). Differential abundance with ALDEx2. Statistical analyses: normality by Shapiro–Wilk; group comparisons by t-test or Mann–Whitney as appropriate; Kruskal–Wallis with Benjamini–Hochberg for alpha diversity; PERMANOVA for beta diversity. Spearman correlations for associations with dam/sire/puppy parameters. Machine learning: Supervised model with leave-one-out cross-validation using rectal microbiome features and activity score to predict FPS (death). Day 1 test cohort: 40 puppies (27 healthy, 13 FPS); Day 8: 28 (20 healthy, 8 FPS). Model performance reported via Precision-Recall AUC and top predictive taxa identified.

• Cohort: 165 puppies from 25 litters; breeds small (n=48), medium (n=81), large (n=36). Median litter size 6; median dam age 4 years.
• FPS occurrence: 22/165 puppies (13.3%) from 12/25 litters (48%); mortality 100% (death 2–11 days PP; median 3.5).
• Risk associations: FPS not associated with dam age, parity, breed, season, kennel, parturition abnormalities, litter size, sex, or rectal temperature. FPS occurred more frequently in puppies with low Day-1 activity scores (P<0.001). FPS in one puppy did not increase risk in littermates.
• Necropsy (9/22 FPS): 6 with diffusely hyperemic intestinal mucosa; 1 with congested, hyperemic caudal lung lobes; 2 without gross abnormalities.
• Community composition: Proteobacteria and Firmicutes dominated Day 1 and Day 8 rectal microbiomes in both groups; other phyla included Bacteroidetes, Fusobacteria, Verrucomicrobia, Actinobacteria, Deferribacteres, and Tenericutes.
• Temporal differences within groups: In healthy puppies, higher Day-1 relative abundance of Epulopiscium and Clostridium celatum and Clostridium perfringens compared to Day-8. In FPS puppies, Streptococcus was lower on Day-1 vs Day-8.
• Diversity metrics: Within-group alpha diversity did not differ between Day-1 and Day-8 in healthy (P=0.33) or FPS (P=0.22) puppies; overall low alpha diversity with high inter-individual variability. Beta diversity differed between Day-1 and Day-8 for both groups (P=0.001).
• Between-group differences: Alpha diversity did not differ between healthy and FPS on Day-1 (P=0.96) or Day-8 (P=0.14). Beta diversity differed between healthy and FPS on Day-1 (P=0.003) and Day-8 (P=0.005), indicating distinct community structures.
• Proteobacteria/Firmicutes ratio: Day-8 ratio higher in FPS vs healthy (P=0.013); Day-1 not different (P=0.86). Relative decreased Clostridia and Enterococcus and increased Pasteurellaceae associated with FPS.
• Predictive modeling: Day-1 rectal microbiome composition (with activity score) accurately predicted FPS-related death (Precision-Recall AUC = 0.8). Day-8 microbiome was not predictive. Pasteurellaceae featured among top 10 predictive taxa.
• No significant correlations between rectal bacterial diversity and dam/puppy parameters (e.g., dam age/body condition, puppy rectal temperature, body weight).

The study demonstrates that neonatal puppies which later succumb to FPS already harbor distinct rectal microbiome configurations by Day 1 post-partum, evidenced by significant beta diversity differences and characteristic taxonomic shifts. These include increased Pasteurellaceae and a higher Proteobacteria/Firmicutes ratio by Day 8, alongside reduced Clostridia and Enterococcus in FPS puppies. Such patterns suggest early dysbiosis that may impair colonization resistance and mucosal development, potentially facilitating pathogen overgrowth and systemic disease. The findings parallel dysbiosis signatures seen in human neonatal diseases such as necrotizing enterocolitis, supporting a possible cross-species pathophysiologic theme. Importantly, Day-1 microbiome features were predictive of FPS-related mortality, indicating potential for early risk stratification and targeted monitoring. The enrichment of possibly orally derived taxa (e.g., Pasteurellaceae) in FPS puppies may reflect aberrant acquisition or compromised upper gastrointestinal defenses, warranting further investigation. Conversely, higher Clostridia in healthy puppies may confer benefits via butyrate production and colonization resistance, suggesting potential probiotic or microbiome-restorative strategies. Overall, the results implicate early-life microbiome composition as a significant factor in FPS and a promising biomarker and therapeutic target.

FPS is common and highly fatal in purebred dogs. Distinct rectal microbial phenotypes are present as early as Day 1 post-partum in puppies that later develop FPS, characterized by altered beta diversity, reduced Clostridia and Enterococcus, increased Pasteurellaceae, and a higher Proteobacteria/Firmicutes ratio by Day 8. Day-1 microbiome composition predicts FPS-related death, highlighting the potential for microbiome-based early diagnostics and interventions to reduce mortality. Future research should include larger, longitudinal cohorts integrating dam (oral, intestinal, vaginal, milk, skin) and puppy microbiomes and metabolomes, comprehensive necropsy and pathogen assessments, and evaluation of microbiome-modulating therapies. Similarities to human neonatal dysbiosis-associated diseases suggest one-health relevance and shared therapeutic opportunities.

• Sample size modest, potentially limiting statistical power, though findings align with prior literature and remained robust despite litter/breed/kennel signatures.
• Maternal microbiomes (oral, intestinal, vaginal, milk, skin) were not sampled, limiting assessment of transmission sources and maternal influence.
• Use of rectal swabs may introduce contamination from perianal skin or dam licking; however, similar maternal care across litters reduces systematic bias.
• Comprehensive necropsies, tissue cultures, and fecal pathogen testing were not routinely performed in deceased puppies, limiting determination of specific causes of death.
• Rarefaction depth constrained by low DNA concentrations may underrepresent low-abundance taxa.