Microbial, metabolic and molecular determinants of neonatal mortality in the rural areas of Limpopo Province, South Africa: umblical cord blood metabolomic profile and bacterial anaysis and their association with perinatal complications

Abstract

BACKGROUND: Neonatal mortality continues to pose a substantial public health challenge globally, particularly in developing countries. Some primary contributors are microbial infections, adverse pregnancy and birth outcomes (APBOs), and labor or obstetric complications. Despite this recognition, the underlying molecular mechanisms of these infections and perinatal complications are poorly understood. Interestingly, microbial and metabolomic approaches have emerged as effective techniques for the early detection and identification of biomarkers for microbial infections and perinatal complications, shedding more light on understanding the underlying molecular mechanisms of these adverse outcomes. Hence, this study employed microbial analysis (through 16S rRNA PCR assay and antibiotic susceptibility test profiling) and metabolomic profiling of umbilical cord blood to identify potential biomarkers for perinatal complications. METHODS: A cross-sectional study utilizing purposive sampling was conducted at selected district hospitals in the Vhembe district from May 2023 to September 2023. Following participants' consent, questionnaires were used for data collection and umbilical cord blood samples were collected from 129 participants. Blood culture was done, and antibiotic susceptibility testing was conducted on the isolates. Additionally, genomic DNA was isolated from the blood samples for the detection of pathogenic bacterial strains, using a conventional PCR assay. Untargeted metabolomic profiling approach using Liquid Chromatography-Quadrupole-Time-of-flight Mass Spectrometry (LC-q-tof-MS) was conducted. The data were entered into REDCap and exported to IBM Statistical Package for the Social Sciences (SPSS) software, version 26, for analysis. Descriptive analysis and multivariate logistic regression analysis were conducted, with statistical significance set at a p-value less than 0.05 at a 95% confidence level. For metabolomic analysis, multivariate analysis approaches were employed to identify signals that differed between perinatal complications and uncomplicated pregnancies. Metabolic pathway analysis of perturbed metabolites was conducted using MetaboAnalyst 6.0. RESULTS: Among 129 participants, the overall occurrence of perinatal complications stood at 35/129 (27.13%). The majority of participants, accounting for 55/129 (42.64%), fell within the age range of 20 to 29 years, while a significant portion (83.7%), were unemployed. Factors such as maternal age, birth weight, maternal blood pressure, anesthesia use, and delivery mode showed associations with perinatal complication risk, with corresponding P-values of 0.032, 0.027, 0.041, <0.001, and <0.001, respectively. Additionally, bloodstream infection (bacteremia) prevalence, as detected by culture-dependent and 16S rRNA PCR assays, was recorded at 30.23% and 26.36%, respectively. The majority of culture-confirmed bacterial isolates (58.06%) were gram-positive, with S. epidermidis (36.73%) and S. aureus (20.43%) being the predominant strains. Notably, E. coli infections showed a significant association with perinatal complications (P = 0.001). Bloodstream infection (Bacteremia) correlated significantly with maternal educational level, maternal blood pressure at birth, gestational booking stage, and pre-pregnancy BMI. Most culture-confirmed isolates exhibited high levels of antibiotic resistance to ampicillin, ceftazidime, and cefoxitin, while gentamicin, imipenem, amikacin, and ciprofloxacin proved effective against both gram-positive and gram-negative isolates. Furthermore, 93.38% of the tested isolates displayed multidrug resistance (MDR). Through untargeted metabolomic profiling analysis and multivariate analysis using the orthogonal partial least squares discriminant analysis (OPLS-DA) model, 107 metabolites were identified, showcasing differences between perinatal complications and uncomplicated pregnancies. Among the 107 perturbed metabolites, univariate analysis highlighted 50 upregulated and 57 downregulated metabolites in perinatal complications at P<0.05. Furthermore, the affected metabolic pathways, including ethyl lipid metabolism, sphingolipid metabolism, and glycerophospholipid metabolism, were identified as statistically significant. CONCLUSION: Our study revealed a high occurrence rate of perinatal complications (25.58%), with maternal high blood pressure, maternal age (10-19 years), low birth weight, delivery mode, and anesthesia for C-sections identified as significant contributors. Proactive healthcare interventions during antenatal care visits are crucial to minimize complications. Additionally, when looking at microbial analysis, there was no significant association between bacteremia and perinatal complications but highlighted a higher prevalence of bloodstream infections, linked to factors like maternal education level and BMI. Furthermore, upon conducting metabolic profiling, it was evident that specific umbilical cord blood processes were closely associated with perinatal complications, indicating their potential as biomarkers for assessment, prediction, and early intervention strategies.