Abstract
BACKGROUND: Nonalcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease. Particulate matter air pollution <2.5 μm in diameter (PM(2.5)) is a ubiquitous exposure primarily produced from fossil fuel combustion. Previous epidemiologic studies have been mixed. The objective of this study was to examine the association between ambient PM(2.5) exposure and NAFLD among hospitalized patients in the Nationwide Inpatient Sample (NIS). METHODS: We conducted a cross-sectional analysis of hospitalizations from 2001 to 2011 using the NIS, the largest nationally representative all-payer inpatient care administrative database in the United States. Average annual PM(2.5) exposure was estimated by linking census tracts (based on NIS-provided hospital ZIP Codes) with a spatiotemporal exposure model. Clinical conditions were identified using hospital discharge diagnosis codes. Multivariable logistic regression incorporating discharge weights was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for the association between PM(2.5) exposure and odds of NAFLD among hospitalized patients adjusting for age, sex, race/ethnicity, year, individual- and area-level socioeconomic status, urbanicity, region, obesity, diabetes, metabolic syndrome, impaired fasting glucose, dyslipidemia, hypertension, obstructive sleep apnea, and smoking. RESULTS: There were 269,705 hospitalized patients with NAFLD from 2001 to 2011 (total unweighted n = 45,433,392 hospitalizations). Higher ambient PM(2.5) exposure was associated with increased odds of NAFLD among hospitalized patients (adjusted OR: 1.24 per 10 μg/m(3) increase, 95% CI 1.15-1.33, p < 0.01). There were statistically significant interactions between PM(2.5) exposure and age, race/ethnicity, diabetes, smoking, and region, with stronger positive associations among patients who were aged ≥45 years, non-Hispanic White or Asian/Pacific Islander, non-diabetics, non-smokers, or in the Midwest and West regions, respectively. CONCLUSIONS: In this nationwide cross-sectional analysis of the NIS database, there was a positive association between ambient PM(2.5) exposure and odds of NAFLD among hospitalized patients. Future research should examine the effects of long-term historical PM(2.5) exposure and incident NAFLD cases.