Objective To investigate the prevalence,natural history,and causes of death of nonalcoholic fatty liver disease( NAFLD),as well as related influencing factors. Methods A total of 833 retired cadres and staff members who underwent physical examination in Shanghai Changzheng Hospital and Shanghai 85 Hospital of the PLA from January 1 to December 31,2011 and received follow- up visits in either hospital every year since 2011 were enrolled as study subjects,and were divided into NAFLD group( 459 patients who were diagnosed with NAFLD before December 31,2011) and control group( 374 patients without liver or biliary diseases). The patients' clinical data were collected,including body height,body weight,systolic pressure,diastolic pressure,blood biochemical parameters,presence or absence of diabetes,hyperlipidemia,cerebrovascular and cardiovascular diseases,and malignant tumor,and smoking and drinking,and the death time and causes of death were clarified for the patients who died. The prevalence and natural course of NAFLD and related risk factors and prognostic factors were analyzed in this population. The t- test was applied for comparison of continuous data between groups,the chi- square test was applied for comparison of categorical data between groups,the multivariate binary logistic regression was applied to analyze the risk factors for the pathogenesis of NAFLD,and the multinomial logistic regression was applied to analyze the influencing factors for aggravation or alleviation of NAFLD. Results The patients in NAFLD group accounted for 55. 1% of all subjects,and the proportion of male patients was higher than that of female patients( 58. 0% vs 46. 7%,χ2= 4. 962,P = 0. 026). Compared with the control group,the NAFLD group had significantly higher body mass index( BMI),systolic pressure,diastolic pressure,alanine aminotransferase( ALT),fasting blood glucose,serum uric acid,and triglyceride( TG),a significantly higher proportion of male patients,and significantly higher proportions of patients with hypertension,diabetes,metabolic syndrome,newly- onset diabetes,and newly- onset metabolic syndrome( all P < 0. 05),as well as a significantly lower high- density lipoprotein cholesterol level( t =- 9. 818,P < 0. 001). Of all 374 patients in the control group,64( 17. 1%) had newly- onset NAFLD,and the incidence rate of newly- onset NAFLD showed no significant difference between male and female patients( 18% vs 15%,χ2= 0. 349,P > 0. 05); BMI and TG were the independent risk factors for newly- onset NAFLD( OR =1. 304,95% CI: 1. 153- 1. 474,P < 0. 001; OR = 1. 509,95% CI: 1. 080- 2. 166,P = 0. 016). At the end of follow- up in NAFLD group,235 patients( 51. 2%) achieved a stable disease,40( 8. 7%) achieved improvement,144( 31. 4%) achieved remission,and 40( 8. 7%) experienced aggravation. The increase in BMI increased the risk of aggravation of NAFLD( OR = 1. 163,95% CI: 1. 028- 1. 316,P = 0. 016),while reductions in BMI and serum uric acid were the protective factors against NAFLD( OR = 0. 849,95% CI: 0. 781-0. 923,P < 0. 001; OR = 0. 997,95% CI: 0. 995- 1. 000,P = 0. 038). Ten patients in NAFLD group died( including 5 patients with malignant tumor and 2 patients with cerebrovascular and cardiovascular diseases),and 6 patients in the control group died( including 2 patients with malignant tumor,no patients with cerebrovascular and cardiovascular diseases),and the risk of death showed no significant difference between the two groups( P > 0. 05). Conclusion The prognosis of NAFLD is relatively good in people who underwent physical examination,and BMI,TG,and serum uric acid are the factors affecting the prognosis of such patients. The major causes of death in NAFLD patients are malignant tumor and cerebrovascular and cardiovascular diseases,rather than end- stage liver disease.