Associations Of Serum Uric Acid, Bilirubin And Their Change With Risk Of Incident Cardiovascular Disease In A Prospective Cohort Study

Cardiovascular disease(CVD)has become the leading death cause,accounting for more than 40% of the deaths of Chinese residents.To prevent and treat CVD has become an urgent public problem.The pathological mechanisms of CVD are complex,it's believed that oxidative stress and inflammation play a key role in the development of CVD.Serum uric acid(SUA)is the end product of purine metabolism,though SUA itself has antioxidant properties,high SUA mainly play a role in promoting oxidative stress,inducing inflammation and leading to endothelial dysfunction in cells.Whereas bilirubin is a physiological antioxidant of heme metabolism,it may have a protective effect on CVD,but low levels of bilirubin may also cause lipid peroxidation and lead to hardening of blood vessels.The two-way effect of SUA and bilirubin may be related to the serum levels themselves.In addition to being regulated by genetic variation,the serum levels of SUA and bilirubin are related to various lifestyles such as smoking,drinking and exercise,however,few studies have explored the effect of green tea,night sleep,nap,and occupational factor on SUA and bilirubin levels.Previous epidemiology studies on the associations of SUA and bilirubin with the risk for incident CVD have not yet reached consistent results,so SUA or bilirubin has not been identified as an independent risk factor for CVD.Simultaneously,most Mendelian Randomization(MR)studies were conducted in European and American populations and have failed to confirm whether there are causal associations between SUA,bilirubin and incident CVD risk,and there is little evidence in Chinese.In addition,the latest cross-sectional studies suggests that the copresene of high SUA and low bilirubin level have potential additive interactions on the increased risk of incident CVD in the diabetic population,but no prospective studies have reported the associations of SUA and bilirubin change with the risk of incident CVD in the general population.In summary,the current study aimed to explore the effects of major lifestyle and shift work on SUA and bilirubin,and to clarify the independent and combined effect of SUA and bilirubin and their 5-year changes on the risk of CVD incidence and whether causal associations exist between SUA and bilirubin and CVD in the Dongfeng-tongji cohort.The study will provide scientific evidence for taking measures such as changing lifestyles to control SUA and bilirubin levels and further early detecting and intervening high risk CVD population and CVD patients.The main contents are as follows:Part 1 Effects of major lifestyle and occupational factor on serum uric acid and bilirubinObjective: To explore the relationship of major lifestyles and occupational factor with 5-year changes in SUA and bilirubin levels and the risk of hyperuricemia and hyperbilirubinemia incidence.Methods: We included 14607 participants who were free of self-reported or diagnosed CVD,cancer,renal diseases,estimated glomerular filtration rate(e GFR)< 15 m L/min/1.73m2 at baseline and loss of follow-up.Multivariate generalized linear regression and logistic regression models were used to analyze the associations of major lifestyle and shift work with 5-year changes in SUA and bilirubin levels and risk of hyperuricemia and hyperbilirubinemia incidence.Results: For SUA,compared with those with education level of primary school or below,the average SUA level in individuals with education level at high school or above decreased by 9.86 ?mol/L,and the risk of hyperuricemia incidence decreased 11%;similarly,the average SUA level decreased 4.80 ?mol/L in green tea drinkers,compared with non-tea drinkers.Compared with subjects of napping within 30 min,the SUA level in those with napping at 30-60 min decreased 4.38 ?mol/L.On the contrary,the risk of hyperuricemia in current smokers increased 25%,compared to non-smokers.The SUA level and risk of hyperuricemia in current drinkers increased 4.45 ?mol/L and 28%,respectively,compared to non-drinkers.In addition,compared with subjects of non-shfit work,the average SUA levels increased 4.83 ?mol/L and 6.20 ?mol/L respectively after 10-20 years and over 20 years of shift work,which showed a dose-response association,however,shift work years were not significant with the risk of incident hyperuricemia.No associations were found between physical activity,night sleep and risk of hyperuricemia.For bilirubin,compared with subjects with education level of primary school or below,the risk of hyperbilirubinemia in those with education level of high school or above decreased 22%.Similarly,the risk of hyperbilirubinemia in current smokers decreased 19%,compared to non-smokers.Compared with night sleep at 7-8 h and nap within 30 min,respectively,the risk of hyperbilirubinemia in subjects with night sleep over 9 h and no-napping decreased 24% and 15%.To the opposite,compared with nondrinkers,the risk of hyperbilirubinemia in current drinkers increased 21%.Shift work years were associated with 5-year change in bilirubin in a dose-response manner.Compared with non-shift workers,the bilirubin level increased 0.47 ?mol/L and the risk of incident hyperbilirubinemia increased 17%.No association was found between drinking green tea,physical activity and the risk of incident hyperbilirubinemia.Conclusions: Education level,drinking green tea,smoking,drinking,nap and shift work years were related to SUA and bilirubin level and risk of incident hyperuricemia and hyperbilirubinemia,which indicated that changing lifestyle might contribute to control SUA and bilirubin levels.Part 2 Associations of serum uric acid level at baseline and its change with risk of incident cardiovascular diseaseObjective: To study the relationship of SUA at baseline and its 5-year change with risk of incident CHD,stroke and its subtyper and CVD in a prospective study and to further explore whether the above association was causal in Mendelian randomization(MR)study.Methods: We included 16419 participants who were free of self-reported or diagnosed CVD,cancer,renal diseases,e GFR<15 m L/min/1.73m2 and missing values of SUA or other covariates at baseline and were followed until Dec,31 2016.Cox regression was used to evaluate multivariable adjusted hazard ratio(HR)and confidence interval(95% CI).After further excluding incident CVD before Dec.31,2013,subjects with e GFR<15 m L/min/1.73m2 and missing values of SUA at first follow-up in 2013,we totally included 11777 subjects who were followed until Dec.31,2016 to study the association of 5-year change in SUA with risk of incident CVD.Participants were divided into four categories according to 5-year change of hyperuricemia between baseline and first follow-up: non hyperuricemia,remittent hyperuricemia,incident hyperuricemia,and persistent hyperuricemia.Simultaneously,we also included 6072 subjects to conduct a MR analysis.7 single nucleotide polymorphism(SNPs)related to SUA were selected as genetic instrument variable to determine whether a causal association existed between SUA and the risk of incident CVD.Results: Compared with the lowest quartile of SUA at baseline,the adjusted HR(95% CI)of CHD incidence across sex-specific quartiles were 1.18(1.05,1.31),1.16(1.04,1.30),1.31(1.17,1.46),Ptrend<0.001,SUA was related to the risk of CHD incidence in a dose-response manner.The above association was more evident in age ? 65 years,female,non-overweight,normotensive,non-diabetes,non-hyperlipidemia and individuals without metabolic syndrome or with normal renal function.Similar association was observed between SUA and CVD.However,no such association exited between SUA and the risk of stroke with its subtype incidence.Compared with non-hyperuricemia both at baseline and follow-up,the risk for incident CHD,stroke and CVD in participants with incident hyperuricemia increased 10%,6% and 6%(all P>0.05),respectively;the risk for incident CHD,stroke and CVD in participants with persistent hyperuricemia increased 29%,59% and 32%(all P<0.05),respectively.The MR study failed to confirm a causal association between SUA and the risk of CVD incidence.Conclusions: SUA levels was related to the risk of incident CHD in a dose-response manner.5-year persistent hyperuricemia both at baseline and follow-up was associated with increased risk of incident CHD,stroke,and CVD.No causal association between SUA and CVD risk was found through MR analysis.Part 3 Associations of serum bilirubin level at baseline and its change with risk of incident cardiovascular diseaseObjective: To study the relationship of three types of bilirubin at baseline and its 5 –year change with risk of CHD,stroke and CVD incidence in a prospective study and to further explore whether the above association was causal in MR study.Methods: We included 12644 participants who were free of self-reported or diagnosed CVD,cancer,and potential liver disease,biliary disease,and renal diseases at baseline.Cox regression was used to evaluate multivariable adjusted HR and 95% CI.After further excluding incident CVD before Dec.31,2013 and missing values of bilirubin at first follow-up in 2013,we totally included 9721 subjects who were followed until Dec.31,2016 to study the association of 5-year change in bilirubin with risk of incident CVD.Participants were defined as low or high bilirubin group for direct bilirubin(DBIL),indirect bilirubin(IBIL)and total bilirubin(TBIL),respectively.We defined 4 categories according to bilirubin 5-year change between baseline and follow-up: persistent low level group,remittent low level group,incident high level group,persistent high level group.Simultaneously,we included 6024 individuals for MR analysis.5 SNPs related to bilirubin were selected as genetic instrument variable to determine whether a causal association existed between bilirubin and the risk of CVD incidence.Results: DBIL was associated with CHD,stroke and CVD in a positive dose-response manner.Compared with the lowest quintile of DBIL at baseline,the adjusted HR(95% CI)of CHD incidence across quintile 2 to quintile 5 were 1.34(1.16,1.54),1.31(1.14,1.51),1.34(1.17,1.54),1.41(1.22,1.63),Ptrend<0.001;the adjusted HR(95% CI)of stroke incidence across quintile 2 to quintile 5 were 1.44(1.11,1.87),1.58(1.23,2.03),1.67(1.30,2.14),1.65(1.28,2.13),Ptrend<0.001.The similar association was observed between DBIL and the risk of incident CVD.There was an L-shaped association between IBIL and TBIL at baseline and the risk of incident CHD or CVD.With the first quintile of IBIL at baseline as reference,the subjects with IBIL level at quintile 4 have the lower risk of CHD(HR=0.87,95% CI: 0.76,0.96)and CVD(HR=0.89,95% CI: 0.79,0.98)respectively.Similarly,the participants with TBIL at quintile 3 have the lowest risk for CHD(HR=0.85,95% CI: 0.74,0.97)and CVD(HR=0.88,95% CI: 0.78,0.99).However,no significant association was found between IBIL,TBIL and the risk of stroke incidence.Compared with persistent low level of DBIL,IBIL,and TBIL both at baseline and follow-up respectively,persistent high level of DBIL,IBIL and TBIL were not statistically related to the risk of CHD,stroke and CVD incidence.The MR study failed to confirm a causal association between TBIL and the risk of incident CVD.Conclusions: Elevated DBIL was associated with increased risk of incident CHD,stroke and CVD in a dose-response manner,while there was an L-shaped relationship of IBIL and TBIL with the risk of CHD and CVD incidence,which implicated that mildto-modest increased IBIL and TBIL might decrease the risk of CHD and CVD incidence.No causal association between TBIL and the risk of incident CVD was found through MR analysis.Part 4 Combined effect of serum uric acid and bilirubin level for risk of incident cardiovascular and cerebrovascular diseaseObjective: To explore the combined effect of SUA and bilirubin level on the risk of incident CVD in a prospective study.Methods: We included 12,643 participants who were free of CVD,cancer,potential liver disease,biliary disease,renal diseases and missing values of SUA,bilirubin and other covariates at baseline.Participants were defined as low or high-level group and then were combined into 4 categories: low SUA level and low bilirubin level,low SUA level and high bilirubin level,high SUA level and low bilirubin level,and high SUA level and high bilirubin level.Because high SUA level,DBIL and low IBIL and TBIL level were related to increased risk of CVD incidence,we set low SUA level and low DBIL level group and low SUA level and high IBIL(TBIL)level as reference group respectively in the analyses of combined effect of SUA with DBIL,IBIL and TBIL on risk of CVD incidence.Cox regression was used to evaluate multivariable adjusted HR and 95% CI.Results: Compared with low SUA level and low DBIL level group,the increased risk for CHD,stroke and CVD in participants with high SUA level and high DBIL level group were 24%,49%,and 27%(all P<0.05),respectively.On the contrary,compared with low SUA level and high IBIL level group,the increased risk for CHD,stroke and CVD incidence in participants with high SUA level and low IBIL level group were 17%,23%,and 16%(all P<0.05),respectively.The combined effect of SUA and TBIL was similar with IBIL.Conclusions: The copresence of high level SUA with high DBIL,low IBIL or TBIL respectively had combined effect on increased risk for CHD,stroke and CVD incidence.