Application of multi-parameter population model based on endogenous population biomarkers and flow volume in wastewater epidemiology

Highlights: • 4-PA and MIAA were firstly selected as WBE-population biomarkers in China. • Population biomarkers and flow volume based population model were established. • Consumption of PPCPs was estimated by WBE based on multi-parameter population model. The consumption or prevalence of acesulfame, caffeine, paracetamol and amantadine was estimated by wastewater-based epidemiology based on a multi-parameter population model in 20 sewage treatment plants (STPs) in Hebei province, China. To minimize the uncertainties contributed by population estimation in WBE, a multi-parameter population model was established based on the population biomarkers equivalent population and flow volume-population with the weight factors calculated by the analytic hierarchy process (AHP). 4-Pyridoxic acid (4-PA), cotinine, trans-3′-hydroxycotinine (trans-3′-OH-Cot) and 1,4-methylimidazole acetic acid (MIAA) were selected as population biomarkers. The estimated model population showed the highest correlations (r² = 0.97, p < 0.01) and lowest variation (one way-ANOVA, p = 0.82, mean variation: −0.1%) comparing to the census data, suggestion better population estimation. The estimated consumption of acesulfame, caffeine, paracetamol and amantadine was 6.7 ± 2.4 mg/day/inh, 50.5 ± 38.5 mg/day/inh, 61.5 ± 52.7 mg/day/inh and 0.52 ± 0.33 mg/day/inh, respectively. Meanwhile, the prevalence of paracetamol and amantadine was calculated to be 5.3% ± 4.5% and 0.28% ± 0.18%, respectively. The estimated results were consistent with that of previous researches in China and were also in accordance with the consumption calculated by sales data (acesulfame and paracetamol). Moreover, uncertainty study showed decrease in population-associated uncertainties by using a multi-parameter population model. The results demonstrated that the multi-parameter population model constructed in this research is feasible to apply in WBE and might lead to lower uncertainties in population estimation.