| Unhealthy behavioral and metabolic factors are major contributors to mortality and disease burden worldwide.An increasing number of cohort studies,which constructed lifestyle scores using multiple behaviors to reflect individuals’ overall lifestyles,have investigated the associations between lifestyle scores and mortality.However,cohort studies from China or comparing the associations across countries were limited.Besides,few studies investigated the modification of metabolic status on the association between combined lifestyle factors and mortality,which would facilitate the health behavior management among individuals with different metabolic health levels given the interrelation of behavioral and metabolic factors.Socioeconomic status is the upstream determinant of behavioral factors,and the behavioral patterns among populations with different socioeconomic backgrounds could be varied.Thus,it is necessary to investigate the modification of socioeconomic status on the association between combined lifestyle factors and mortality,which could facilitate lifestyle management among populations with different socioeconomic backgrounds;however,evidence is sparse and inconsistent.Besides,previous studies only focused on single socioeconomic factors(e.g.,income,education,or occupation).On the other hand,to further understand the associations of combined lifestyles and socioeconomic status with mortality,it is inevitable to investigate the mediation of lifestyles on the association between socioeconomic status and mortality,which could determine whether lifestyle intervention could narrow the health disparity between populations with different socioeconomic backgrounds;however,current evidence is heterogeneous.Besides,previous studies only considered single socioeconomic factors and limited behavioral factors,which could not comprehensively reflect individuals’ socioeconomic status and lifestyles.Hence,in cohorts from China,the US,and the UK,healthy lifestyle scores and metabolic health scores were constructed to reflect individuals’ overall health behaviors and overall metabolic status,respectively,and latent variables reflecting the overall socioeconomic status were constructed.The study investigated the associations between healthy lifestyle scores and mortality in the whole population and populations with different metabolic health and socioeconomic levels,as well as the mediation of healthy lifestyles on the association between socioeconomic status and mortality.The study included five cohorts.The China Health and Nutrition Survey(CHNS)recruited 23311 participants from 12 provinces in China between 1997 and 2011.The Dongfeng-Tongji Cohort(DFTJ)recruited 41129 retirees from the Dongfeng Motor Corporation in Hubei province between 2008 and 2013.The Kailuan Study recruited 159017 employees and retirees from the Kailuan Group in Hebei province between 2006 and 2013.The US National Health and Nutrition Examination Survey(NHANES)selected a series of nationally representative samples of 54267 individuals in 1988~2014.The UK Biobank(UKB)recruited 502491 participants from 22 centers in the UK between 2006 and 2010.Information on demographic and socioeconomic characteristics,lifestyles,disease history,physical examination,and laboratory tests were collected according to standard procedures.CHNS collected death information through household follow-ups,and other cohorts collected death information through linkages to registry systems.Part 1 Association between combined healthy lifestyles and mortalityObjective: To investigate the associations between combined healthy lifestyles and mortality among adults from China,the US,and the UK.Methods: After excluding those with missing information on lifestyles,covariates,and vital status,708263 participants from the CHNS,DFTJ,Kailuan Study,NHANES,and UKB were included in the analyses.According to previous publications and the recommendations from the World Health Organization,healthy lifestyle scores were constructed using the information on smoking,alcohol drinking,leisure-time physical activity,and diet.Each factor was divided into healthy and unhealthy levels,and 1 and 0points were assigned for the healthy and unhealthy levels of each factor,respectively.The healthy lifestyle scores were the sum of the points and ranged between 0 and 4,with higher values indicating healthier lifestyles.Since the numbers of participants with healthy lifestyle scores of 0 and 4 points were small,participants were regrouped into three groups according to healthy lifestyle scores(i.e.,0~1,2,and 3~4 points)to increase the statistical power.In each cohort,the Cox regression model adjusted for baseline demographic characteristics,socioeconomic features,body mass index,and major noncommunicable disease histories was used to evaluate the association between the healthy lifestyle score and mortality.Hazard ratios from three Chinese cohorts were pooled by the random-effects models of meta-analyses.Results: There were 55115 deaths(19751 from three Chinese cohorts,9052 from the NHANES,and 26312 from the UKB)during 7727959 person-year follow-ups.Hazard ratios(95% confidence intervals)for all-cause mortality comparing participants with lifestyle scores of 3~4 points versus 0~1 point were 0.69(0.59~0.82)in Chinese cohorts,0.61(0.55~0.66)in the US NHANES,and 0.63(0.60~0.65)in the UKB.Conclusions: Combined healthy lifestyles were associated with lower mortality among adults from China,the US,and the UK.Part 2 Modification of metabolic health on association between combined lifestyles and mortalityObjective: To investigate the associations between metabolic health and mortality among adults from China,the US,and the UK,and to investigate the modification of metabolic health on the association between combined lifestyles and mortality.Methods: After excluding participants with missing information on metabolic metrics,34350,146561,41873,and 409774 participants from the DFTJ,Kailuan Study,NHANES,and UKB were included in the analysis,respectively.Participants from the CHNS were excluded from the analysis since CHNS did not collect participants’ blood samples in 1997~2006 and could not derive the metabolic health score.According to the National Cholesterol Education Program Adult Treatment Panel III,metabolic health scores were the number of healthy metrics,i.e.,no central obesity,normal blood high-density lipoprotein cholesterol level,normal blood pressure level,and normal blood glucose level.Participants were divided into three groups according to metabolic health scores(i.e.,0~1,2,and 3~4 points)to increase the statistical power.In each cohort,Cox regression models were used to investigate(1)the association between the metabolic health score and mortality and(2)the associations of healthy lifestyle scores with mortality in different metabolic health score groups.Hazard ratios from two Chinese cohorts were merged by the random-effects models of meta-analyses.Results: Hazard ratios(95% confidence intervals)comparing participants with metabolic health scores of 3~4 points versus 0~1 point for all-cause mortality were 0.69(0.66~0.72)in Chinese cohorts,0.70(0.64~0.77)in the US NHANES,and 0.62(0.60~0.64)in the UKB.Among participants with different metabolic health scores,higher healthy lifestyle scores were associated with lower mortality.The modification of the metabolic health score on the association between the healthy lifestyle score and mortality was only observed in the Kailuan Study(P for interaction=0.035),and hazard ratios comparing participants with the healthy lifestyle score of 3~4 points versus 0~1 point of all-cause mortality were 0.86(0.76~0.97),0.81(0.74~0.89),and 0.75(0.68~0.82)among individuals with the metabolic health scores of 0~1,2,and 3~4 points,respectively.Conclusions: Meeting more healthy metabolic metrics was associated with lower mortality.Among individuals with different metabolic health levels,adopting combined healthy lifestyles was associated with lower mortality.Part 3 Associations of socioeconomic status and combined lifestyles with mortalityObjective: To investigate(1)the associations between socioeconomic status and mortality among adults,(2)the mediation of combined healthy lifestyles on the association between socioeconomic status and mortality,and(3)the modification of socioeconomic status on the association of combined healthy lifestyles with mortality.Methods: Participants from the DFTJ and Kailuan Study worked in the same companies,respectively,and the socioeconomic variance was too small in these cohorts,which were inappropriate for analysis related to socioeconomic status and were excluded from the analyses.There were 20851 participants from the CHNS(the same as part 1),44935 participants from the NHANES(the same as part 1),and 399250 participants from the UKB(excluding 57690 participants choosing “do not know” or “prefer not to answer”regarding their household income from the participants included in part 1).In each cohort,a latent variable with three classes(i.e.,high,medium,and low socioeconomic status,which reflected different levels of socioeconomic status)was generated by the latent class analysis,according to income,educational level,occupation,and health insurance(the NHANES only).In each cohort,the Cox regression model was used to investigate the association between socioeconomic status and mortality.The difference method was used to calculate the mediating proportion of the healthy lifestyle score.Associations between healthy lifestyle scores and mortality among individuals with different socioeconomic levels were also evaluated.Results: After adjusting for the healthy lifestyle score,demographic characteristics,body mass index,and disease histories,hazard ratios(95% confidence intervals)comparing participants with low versus high socioeconomic status for all-cause mortality were 2.69(1.97~3.69)in the CHNS,2.01(1.80~2.24)in the NHANES,and 1.84(1.75~1.94)in the UKB,and the proportions mediated by the healthy lifestyle score were7.2%(4.8%~10.8%),20.8%(18.5%~23.1%),and 6.3%(5.7%~6.9%),respectively.The associations between higher healthy lifestyle scores and lower mortality were observed in all socioeconomic groups from all cohorts.The modification of socioeconomic status on the association between the healthy lifestyle score and mortality was only observed in the UKB(P for interaction<0.001),and hazard ratios comparing participants with the healthy lifestyle score of 3~4 points versus 0~1 point were 0.82(0.73~0.92)and 0.53(0.49~0.57)among individuals with high and low socioeconomic status,respectively.Conclusions: Disadvantaged socioeconomic status was associated with higher mortality,which was slightly mediated by combined healthy lifestyles.Thus,encouraging populations to adopt healthy lifestyles alone might not substantially reduce the socioeconomic gradient in mortality.However,combined healthy lifestyles were associated with lower mortality in different socioeconomic groups,which highlighted the importance of healthy lifestyles in the prevention of premature deaths across different socioeconomic groups. |
PDF Full Text Request