作者
Ming‐Chieh Li,Chi‐Hsien Chen,Ching-Yu Lin,Tzu‐Chieh Chou,Yue Leon Guo
摘要
BACKGROUND AND AIM Phthalate exposure has been associated with circulating lipid composition. However, limited epidemiological study has been carried out in elderly people. Therefore, we conducted a cross-sectional study to examine the associations between urinary phthalate exposure and lipidomic profile in Taiwanese elderly people. METHODS A cross-sectional survey of the elderly (age >= 65 years) was conducted in Taipei city, New Taipei City, Hualien County, Chiayi County, and Kaohsiung City in Taiwan. A total of 11 urinary phthalates metabolites and 86 lipids (phosphatidylcholines (PCs) and sphingomyelins (SMs)) were assessed. Demographic information was obtained by a structured questionnaire. Serum lipidomic profile was identified using principal component analysis (PCA). A two-step analysis was performed. We first performed multiple linear regressions to evaluate whether urinary phthalate metabolites were related to the PCA-derived components of serum lipid concentrations. Then we performed multiple linear regressions to investigate each of the lipids that were suggestively correlated with some of the phthalates in PCA. Urinary concentrations of phthalate metabolites were log-transformed to normalize distributions. RESULTS A total of 886 participants (58.9% women) were included in the final analysis. There were ten components identified by PCA, which accounts for 63.2% of total variance. After adjusting for age, sex, body mass index, urinary creatinine, urinary phthalate metabolites were correlated with 6 components. Among them, component 7 was correlated with the most phthalates metabolites (Bonferroni corrected p-value < 0.007). Monomethyl phthalate (MMP) and monobutyl phthalate (MnBP) were negatively correlated with component 7, whereas monoisobutyl phthalate (MiBP), monobenzyl phthalate (MBzP), and monoisononyl phthalate (MiNP) were positively correlated with component 7. Among component 7, PC(16:0/16:0), PC(16:0/17:1), PC(16:0/18:1), PC(16:0/19:5), PC(16:0/20:5), PC(16:0/22:4), PC(38:0), PC(P-16:1/22:5), PC(P-40:4), and PC(P-40:5) were highly loaded. CONCLUSIONS Urinary phthalate metabolites were correlated with the lipidomic pattern. Most of them were diacyl-phosphatidylcholines, which might play a role in hepatic lipid metabolism.