| [Background]Bisphenol A(BPA)，the main monomer synthetic resin material, is widely used indaily life. Numerous studies have been shown that BPA is a reproductive toxicity,closely related to human reproductive health, and is highly suspected to be endocrinedisruptors of human beings. At present, there is less studies about BPA and UnexplainedRecurrent Miscarriage (URM). So, it is of great significance to explore the relationshipbetween BPA and URM for human reproductive health. However, accuratedetermination of BPA concentration in the sample is the key to the study.[Objective]1. We took advantage of the existing urine and serum samples of URM patients andcontrols we have collected before this study, to detect the BPA concentration of morningspot urine using high performance liquid chromatographic (HPLC) and liquidchromatographic-tandem mass spectrometry (LC-MS/MS) and serum samples usingHPLC and to explore the correlation between those three methods.2. We expected to compare the difference of serum and urine BPA levels in thecases and the controls, and to further estimate and compare the strength of associationbetween BPA concentration and URM risk.[Methods]1. We conducted a hospital-based1:2matched case-control study in Suzhou City ofJiangsu Province, from October2008to December2011. The cases and the healthypregnant matched controls were recruited from the First Affiliated Hospital of SoochowUniversity, the Second Affiliated Hospital of Soochow University, and the Maternityand Child Care of Kunshan City, respectively. We investigated the demorgraphiccharacteristics, menstrual history, pregnancy history, and conceptive history, etc by in-depth interview using standard-structured questionaire. Finally, there were104URMcases and185controls recruited.10ml morning urine and5-10ml venous blood of casesand controls were collected in the two weeks after being recruited. Due to the differentstudy periods of testing samples and other reasons, the total number of the serum orurine samples after being tested sucessfully is not consistent. Finally, there are69casesand114controls urinary samples detected by HPLC method,62cases and108casesserum samples detected by HPLC method and102cases and162controls urinarysamples detected by LC-MS/MS method, respectively.2. M (Median), mean±standard deviation (X S), and G (Geometric mean) wereused to describe the average BPA concentation. Spearman rank correlation was used toanalyze the correlation between urinary BPA concentration detected by HPLC andLC-MS/MS and the correlation between urinary and serum BPA detected by HPLC.2-test, conditional logistic regression were used to estimate and compare theassociation of the relationship between BPA levels and URM by OR and95%CI. SAS9.2software was performed in all the analysis. A two-sided test P-value≤0.05wasconsidered to be statistically significant.[Results]1. Our results were shown that the HPLC method detecting for urinaryconcentration adjusted by creatinine was significantly correlated with LC-MS/MSmethod (rs=0.2906，P=0.0001). The Serum BPA concentration detected by HPLC wasstatistically significanly associated with the BPA urinary level including BPAHPLC(dectected by HPLC), BPAcrHPLC(dectected by HPLC and adjusted by creatinine),BPALC-MS/MS(dectected by LC-MS/MS) and BPAcrLC-MS/MS(dectected by LC-MS/MSand adjusted by creatinine). The Spearman correlation coefficients were0.1771,0.2471,0.1699and0.2285, respectively (all P values less than0.05).2. The significant difference for the median values of serum BPA in cases(8.67ng/ml) and controls (3.59ng/ml) was found (Z=3.506, P=0.0005). Compared to thegroups with serum BPA levels less than LOD, the women with levels of3.59-12.28ng/ml and12.28ng/ml or above had a significantly higher risk of URM(OR=4.39,95%CI:1.15-16.71and OR=4.95,95%CI:1.77-13.82) after adjustment forage, occupation, education, BMI, passive smoking. The risk of URM was linked withthe increased serum BPA level (P=0.0013for trend test). 3. Median values of non-creatinine-adjusted urinary BPA in cases and controlsdetected by HPLC were95.24ng/ml,41.84ng/ml, respectively, which were transformedas76.57μg/g Cr and22.64μg/g Cr after being adjusted by creatinine levels. AdjustedBPA level was significantly higher in cases than in controls (Wilcoxon test, Z=3.529,P=0.0004). Compared to the group with serum BPA level less than5.84μg/g Cr thewomen with levels of20.87-96.61μg/g Cr and96.61μg/g Cr or above had asignificantly higher risk of URM (OR=7.33,95%CI:1.68-32.01and OR=15.07,95%CI:3.31-68.50) after adjusting for the confounding factors metioned above. The moreserum BPA level was the higher risk of URM was (P=0.0003for trend test).4. For the LC-MS/MS method, we also found that the non-creatinine-adjusted(M=1.66ng/ml) and creatinine-adjusted (M=1.66ng/ml) urinary BPA concentration incase group (0.98μg/g Cr) was significantly higher than controls (0.58ng/ml and,0.40μg/g Cr, respectively. Wilcoxon test, Z=4.476, P<0.0001). Compared to the subjectswith serum BPA level less than0.16μg/g Cr, the women with level of0.40-0.93μg/g Crand0.93μg/g Cr or above had a significantly higher risk of URM (OR=3.91,95%CI:1.23-12.45and OR=9.34,95%CI:3.06-28.44) after adjusting for the confounders.Thehigher urinary BPA concentration was significantly associated with theincreased risk ofURM (P<0.0001for trend test).[Conclusion]1. Our study was shown that using HPLC for testing urine BPA concentration wasweakly correlated with LC-MS/MS method, although correlation has approched thestatistically significant level.2. LC-MS/MS method is one of the better methods to detect the concentration ofBPA in urine and serum.3. Although urinary BPA concentration tested by the HPLC and LC-MS/MSmethods and serum BPA concentration detected by HPLC were largely different, ourepidemiological analysis using the three methods still could obtain a consistentconclusion: serum/urine BPA concentration was significantly associated with theincreased risk of URM, and BPA may play an important role in the URM. |
PDF Full Text Request