Chemometric Analysis and Pattern Recognition Methods Applied to the Authentication of Nigerian Honey Using Multi-Analytical Chemical Profiling

Food fraud in the Nigerian honey market, including adulteration with sugar syrups and mislabelling of geographical origin, imposes economic losses on genuine beekeepers and poses health risks to consumers who use honey medicinally. Existing single-parameter authentication methods are insufficient to reliably detect sophisticated adulteration. This study applied an original multi-analytical chemical profiling approach combined with advanced chemometric pattern recognition to develop an authentication framework for Nigerian honey that can discriminate between genuine monofloral, polyfloral, and adulterated products from different production regions. Honey samples from six geopolitical zones were characterised by twenty-three chemical and physical parameters including nuclear magnetic resonance spectroscopy, melissopalynological pollen analysis, isotope ratio mass spectrometry for carbon, nitrogen, and hydrogen stable isotopes, volatile compound profiling by headspace solid-phase microextraction GC-MS, and phenolic fingerprinting by HPLC-diode array detection. The resulting high-dimensional chemical dataset was subjected to principal component analysis, hierarchical cluster analysis, linear discriminant analysis, partial least squares discriminant analysis, and artificial neural network classification. An original multi-block data fusion strategy that combines low-level raw data fusion with mid-level feature selection was developed and evaluated against single-block models. The multi-block partial least squares discriminant analysis model correctly classified honey samples by geographical zone with 96.4 percent accuracy and by adulteration status with 99.1 percent accuracy in external validation. Carbon stable isotope ratio alone achieved only 82.3 percent adulteration detection accuracy. The chemometric framework developed represents an original contribution to food authentication science and provides NAFDAC with a scientifically validated, multi-parameter honey authentication protocol for enforcement application across Nigeria.

Keywords: honey authentication, chemometrics, multi-analytical profiling, food fraud, stable isotope ratio