The Associations Of Blood Cell Counts Based Chronic Systemic Inflammatory Indices With Total And Cause-specific Mortality In Middle-aged And Older Populations: Results From Two Cohort Studies

Chronic systemic inflammation has been considered as an important risk factor for the development of multiple chronic diseases,including cardiovascular disease（CVD）,cancer,diabetes,chronic kidney disease and neurodegenerative disorders,which collectively account for more than 50%of total deaths.Chronic systemic inflammation always developed in the absence of clinical manifestations.Developing available and reliable chronic systemic inflammatory markers may help identify high-risk populations and guide therapeutic decisions,which is important for reducing burden of chronic diseases.Previous studies reported that abnormalities in white blood cell（WBC）count,WBC subtypes or other blood cell such as platelet count were associated with increased risks of morbidity and mortality.The blood cells played varied roles in chronic systemic inflammation,and interacted with each other.To study the independent and joint effects of these blood cells,neutrophil-to-lymphocyte ratio（NLR）,platelet-to-lymphocyte ratio（PLR）and systemic immune-inflammation index（SII）were proposed to reflect the chronic systemic inflammation.These blood cell counts based chronic systemic inflammatory indices were considered to be novel independent prognostic markers for patients with CVD and certain malignancies.However,there were few studies exploring the associations of these chronic systemic inflammatory indices with total and cause-specific mortality risk in the general population.Objective:To investigate the associations of blood cell counts based chronic systemic inflammatory indices with total and cause-specific mortality risk in the general populations.Methods:Our prospective cohort study were based on the Dongfeng-Tongji cohort（DFTJ cohort）and the National Health and Nutrition Examination Survey 1999-2014（NHANES 1999-2014）.The 27009 participants were recruited during baseline survey in2008,and additional 14120 participants were recruited during the first follow-up in 2013.After excluding individuals without platelet count（n=6728）,and without information of neutrophil or lymphocyte count（n=3880）,the left 30521 participants were included in this study.A total of 82091 participants were recruited in the NHANES 1999-2014.We excluded individuals who did not take health examination（n=1487）,without mortality data（n=38）,without platelet count（n=1236）,and without information of neutrophil or lymphocyte counts（n=82）.Finally,a total of 25,761 participants aged≥40 years old were included in this study.Information on demographic characteristics,lifestyles,and health status were assessed with self-administered questionnaire and interview.The body weight,height,serum biochemical levels and blood cell counts were available in the physical examination.The blood cell counts based chronic systemic inflammatory indices were calculated for each participant as follows:NLR=neutrophil count（×10⁹/L）/lymphocyte count（×10⁹/L）;PLR=platelet count（×10⁹/L）/lymphocyte count（×10⁹/L）;SII=neutrophil count（×10⁹/L）/lymphocyte count（×10⁹/L）×platelet count（×10⁹/L）.In the DFTJ cohort,mortality status was tracked through Dongfeng Motor Corporation’s medical insurance system.The causes of mortality were confirmed by medical records from hospitals and death certificates from Centers for Disease Control and Prevention.The participants were followed up till December,31,2018.In the NHANES 1999-2014,the public-use linked mortality files were released by National Center for Health statistics,including vital status,cause of death,follow-up time from interview date and follow-up time from physical examination date.Mortality data were obtained by matching to the mortality file using exclusive respondent sequence number,and were available till December,31,2015.The outcome included mortality form all-cause,CVD,cancer and other causes.Multivariable linear regression models were conducted to disclose the associations of chronic systemic inflammatory indices level with basic characteristics,including age,sex,race（only in the NHANES 1999-2014）,body mass index（BMI）,smoking status,alcohol drinking status and regular physical activity status.Restricted cubic spline models（RCS）were fitted to show the shapes of the associations between blood cell counts based indices and each outcome.Each of chronic systemic inflammatory index was also divided into three subgroups based on the distribution among survivors in each cohort.Subgroups of each index were classified as follows:low level:<50^th percentile（P50）;middle level:P50～P75;high level:>P75.Hazard ratios（HR）and 95%confidence intervals（CIs）were calculated to examine the associations of subgroups of each chronic systemic inflammatory index with total and cause-specific mortality risk by Cox proportional hazards regression models.Both the RCS models and Cox proportional hazards regression models were adjusted for baseline age（continuous）,sex,race（only in the NHANES 1999-2014）,BMI,education level,smoking status,alcohol drinking status,regular physical activity status and baseline disease histories of hypertension,hyperlipidemia,diabetes,CVD and cancer.Then participants were stratified by sex（males and females）,age（<65 and≥65 years old）,BMI（<25 and≥25 kg/m²）,smoking status（non-smokers,smokers）,alcohol drinking status（non-drinkers,drinkers）and regular physical activity status（physically inactive,physically active）.And the associations between chronic systemic inflammatory indices and total mortality risk were examined in each subgroup.The effect modification was evaluated by likelihood ratio test to compare the models with or without the interaction term between each of these factors and chronic systemic inflammatory index.We further explored the benefit effect of regular physical activity on total and cause-specific mortality risk in each subgroup of chronic systemic inflammatory index.Results:In the DFTJ cohort and NHANES 1999-2014,we found that males tended to have a higher level of NLR than females[β（95%CI）=0.116（0.097,0.136）and 0.092（0.075,0.109）,respectively].Overweight participants had a lower level of NLR than subjects with normal BMI[DFTJ cohort:β（95%CI）=-0.018（-0.033,-0.002）;NHANES1999-2014:β（95%CI）=-0.030（-0.057,-0.003）];the NLR level among underweight and obese subjects were similar to that among subjects with normal BMI（P>0.05）.Physically active participants had a lower NLR level than those physically inactive[DFTJ cohort:β（95%CI）=-0.025（-0.043,-0.007）;NHANES 1999-2014:β（95%CI）=-0.063（-0.083,-0.042）].In both cohorts,males had a lower level of PLR than females[DFTJ cohort:β（95%CI）=-0.095（-0.113,-0.077）;NHANES 1999-2014:β（95%CI）=-0.079（-0.097,-0.062）].Compared to subjects with normal BMI,overweight and obesity participants had a reduced PLR level（P<0.001）.Smokers had a lower PLR level than non-smokers[DFTJ cohort:β（95%CI）=-0.020（-0.039,-0.0003）;NHANES 1999-2014:β（95%CI）=-0.075（-0.094,-0.055）].In both cohorts,age was inversely associated with SII level[DFTJ cohort:β（95%CI）=-0.009（-0.010,-0.007）;NHANES 1999-2014:β（95%CI）=-0.002（-0.003,-0.001）].Smokers tended to have a higher level of SII than non-smokers[DFTJ cohort:β（95%CI）=0.116（0.090,0.142）;NHANES 1999-2014:β（95%CI）=0.051（0.022,0.080）].Physically active subjects had a lower SII than those physically inactive[DFTJ cohort:β（95%CI）=-0.034（-0.057,-0.012）;NHANES 1999-2014:β（95%CI）=-0.092（-0.115,-0.068）].RCS models indicated that in the DFTJ cohort,the risks of total,CVD and other causes mortality increased with the elevated NLR level（all P<0.001）.No significant association between baseline NLR level and cancer mortality risk was observed（P=0.392）.We observed linear associations of PLR level with risk of CVD and cancer mortality（P=0.003 and<0.001,respectively）,and non-significant associations of PLR level with risk of total and other causes mortality（P=0.339 and 0.229,respectively）.As the level of SII increased,the risk of total,CVD and other causes mortality tended to be elevated（all P<0.001）,while the risk of cancer mortality reduced（P=0.009）.When examining the chronic systemic inflammatory indices as categorical variables,compared to participants within low NLR subgroup（<1.8）,those within high NLR subgroup（>2.3）had elevated risks of total,CVD and other causes mortality[HR（95%CI）=1.32（1.22,1.43）,1.61（1.42,1.82）and 1.37（1.18,1.60）,respectively].No significant associations between NLR subgroups and cancer mortality risk were observed（P>0.05）.Compared to subjects in low PLR subgroup（<103）,those in high PLR subgroup（>133）had higher risks of CVD and other causes mortality[HR（95%CI）=1.23（1.07,1.41）and 1.21（1.03,1.43）,respectively];those in middle PLR（103～133）and high PLR subgroups had reduced cancer mortality risk[HR（95%CI）=0.76（0.65,0.89）and 0.86（0.73,1.00）,respectively].Compared to subjects in the low SII subgroup（<332）,those within middle SII（332～465）and high SII（>465）subgroups had increased risk of total mortality[HR（95%CI）=1.12（1.03,1.22）and 1.26（1.16,1.36）,respectively)and CVD mortality[HR（95%CI）=1.36（1.19,1.55）and 1.50（1.32,1.71）,respectively];those within high SII subgroup had elevated other causes mortality risk[HR（95%CI）=1.28（1.09,1.49）].No significant difference of cancer mortality risk was observed between SII subgroups（P>0.05）.In the NHANES 1999-2014,RCS models showed that the risks of total and cause-specific mortality tended to be elevated as the chronic systemic inflammatory indices increased（all P<0.05）,with exception of non-significant association between baseline PLR level and CVD mortality risk（P=0.182）.When examining the chronic systemic inflammatory indices as categorical variables,compared to participants within low NLR subgroup（<2.0）,those within high NLR subgroup（>2.6）had elevated risks of total,CVD,cancer,and other causes mortality[HR（95%CI）=1.49（1.39,1.60）,1.48（1.25,1.76）,1.33（1.13,1.56）and 1.58（1.42,1.75）,respectively].Compared to subjects in low PLR subgroup（<126）,those in high PLR subgroup（>159）had higher risks of total and other causes mortality[HR（95%CI）=1.23（1.13,1.34）and 1.31（1.18,1.46）,respectively];No significant associations of PLR subgroups with CVD and cancer mortality risk were observed（P>0.05）.Compared to those with low SII level（<498）,subjects in high SII subgroup（>691）had higher risks of total,CVD,cancer and other causes mortality[HR（95%CI）=1.38（1.27,1.49）,1.33（1.11,1.59）,1.22（1.04,1.45）and1.47（1.32,1.63）,respectively].In the DFTJ cohort,we did not observe significant interaction of sex,age,BMI groups,smoking status,alcohol drinking status and regular physical activity status with NLR,PLR,SII level on total mortality risk(all P_interaction>0.05).In the NHANES 1999-2014,the associations of high NLR,PLR,SII level with total mortality risk were only significant among physically inactive participants[HR（95%CI）=1.60（1.46,1.74）,1.27（1.16,1.40）and 1.43（1.29,1.58）,respectively];above associations were abolished among the physically active participants[HR（95%CI）=1.12（0.94,1.34）,1.07（0.87,1.31）and 1.14（0.96,1.37）,respectively].Significant or marginal significant interactions of NLR,PLR,SII level with regular physical activity on total mortality risk were observed(P_interaction=0.001,0.098 and 0.040,respectively).In the high NLR,PLR,SII subgroups,compared to physically inactive participants,the physically active subjects had reduced risks of total mortality[HR（95%CI）=0.65（0.57,0.75）,0.68（0.57,0.81）and 0.70（0.60,0.81）,respectively],CVD mortality[HR（95%CI）=0.66（0.54,0.82）,0.58（0.44,0.75）and 0.68（0.54,0.85）,respectively]and other causes mortality[HR（95%CI）=0.53（0.41,0.68）,0.67（0.48,0.93）and 0.62（0.47,0.83）,respectively].No significant association between regular physical activity and cancer mortality risk was observed among participants with high NLR,PLR and SII level.Similar results were observed in the NHANES 1999-2014.In the high NLR,PLR,SII subgroups,regular physical activity was associated with reduced risks of total mortality[HR（95%CI）=0.54（0.46,0.63）,0.61（0.52,0.70）and 0.59（0.51,0.68）,respectively],CVD mortality[HR（95%CI）s=0.59（0.40,0.88）,0.59（0.40,0.86）and 0.41（0.27,0.64）,respectively]and other causes mortality[HR（95%CI）=0.50（0.38,0.56）,0.57（0.45,0.71）and 0.57（0.46,0.70）,respectively].We observed a significant association between regular physical activity and cancer mortality among subjects with high NLR level[HR（95%CI）=0.69（0.50,0.96）],but not among subjects with high levels of PLR and SII（P>0.05）.Conclusions:In both the DFTJ cohort and NHANES 1999-2014,we found that high levels of NLR and SII were associated with elevated risks of total,CVD and other causes mortality,and high PLR level was associated with elevated risk of other causes mortality in the general middle-aged and older populations.In subjects with high levels of NLR,PLR,and SII,regular physical activity could significantly reduce the mortality risks of total,CVD,and other causes.The results provided reliable epidemiological evidence for the effects of blood cell counts based chronic systemic inflammatory indices on total and cause-specific mortality risk in the general populations.In a public health view,regular physical activity was suggested as a promising strategy to attenuate the chronic inflammation and elevated risk of mortality.Further studies are warranted to clarify the possible biology mechanisms underlying the current observations.