Article Text

Health transition of the causes of mortality between 2005 and 2015 in Guangdong, China
  1. Xue-yan Zheng1,
  2. Qian Yi1,2,
  3. Yan-jun Xu1,
  4. Xin-ying Zeng3,
  5. Xiao-jun Xu1,
  6. Ge Chen2,
  7. Shu-li Ma2,
  8. Si-li Tang4,
  9. Li-feng Lin1
  1. 1 Guangdong Provincial Center for Disease Control and Prevention, Guangzhou, China
  2. 2 Department of Epidemiology and Biostatistics, School of Public Health, Guangdong Pharmaceutical University, Guangzhou, China
  3. 3 National Center for Chronic and Non-Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, Beijing, China
  4. 4 School of Public Health, Southern Medical University, Guangzhou, China
  1. Correspondence to Dr Li-feng Lin, Guangdong Center for Disease Control and Prevention, 511430, Guangzhou, Guangdong, China; 1396320174{at}qq.com

Abstract

Objectives To analyse the patterns of transition of health burden for 110 causes of death by stratification of age, sex and geographic regions in Guangdong between 2005 and 2015.

Methods We analysed the age-specific, sex-specific, region-specific mortality in Guangdong based on assembled databases. County-level surveillance data were calculated to inform city-level changes.

Results The age-standardised mortality of all causes, non-communicable diseases (NCDs), communicable diseases, maternal diseases, neonatal diseases, malnutrition and injury declined progressively. Despite declining mortality of NCDs, the overall burden of disease was dominated by NCDs (ie, cerebrovascular disease, chronic obstructive pulmonary disease) that still accounted for 86.93% and 88.12% of death in 2005 and 2015, respectively. Considerable variations across geographic regions were observed (lowest in Pearl River Delta and highest in west Guangdong). There was a modest shift to transport injuries at younger ages and unintentional injuries in the elderly.

Conclusions We have documented a dramatic change in the overall mortality and age-specific, sex-specific and cause-specific mortality in Guangdong province between 2005 and 2015. The significant burden of NCDs remains a major healthcare issue despite the notable progress in reducing mortality in Guangdong, China. Our findings highlight important unmet needs to refine healthcare services by taking into account the inequity of age, sex and geographic regions. Identification of the ‘treatable’ risk factors and improved disease surveillance should be continuously improved to minimised the overall and cause-specific mortality.

  • health policy
  • epidemiology
  • public health

Data availability statement

Data are available upon reasonable request.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

During the past three decades, there have been major changes in the disease profile affecting the health worldwide. The major Global Burden of Disease (GBD) has transitioned from communicable diseases to non-communicable diseases (NCDs), which has been associated with the rapid growth in economics, the marked improvement in healthcare, the increased risk factors related to lifestyle (smoking) and diet, and the accelerating ageing society. Despite profound progress in public healthcare, there remains a considerable burden of disease globally. The global population suffers from numerous healthcare challenges such as emerging infectious diseases, interpersonal conflict and violence, and injury, many of which contributed a significant disease burden and affecting healthcare decision makings.1 Currently, monitoring of the age-specific or sex-specific mortality is indispensable to health surveillance,2 which could inform policy makers of the outcomes of healthcare programmes and the priority for disease prevention.

In support of the health agenda of global sustainable development goals, most GBD studies have provided the comprehensive annualised assessment of multiple underlying causes stratified by age, sex and geographic regions.2–8 According to the 2016 GBD study, the total number of NCDs increased from 2005 to 2015 despite progressively decreased age-standardised rates. Despite minimal changes in the total numbers of mortality associated with injuries, both the total number and age-standardised rates of communicable, maternal, neonatal and nutritional disease (CMNN) mortality decreased, which was accompanied by a decrease in the age-standardised mortality from injuries. In China, ischaemic heart disease (IHD), stroke, chronic obstructive pulmonary disease and cancers (esp. liver, stomach and lung) have accounted considerably for the years-of-life-lost in 2013.6 However, there may be a significant heterogeneity in the cause of mortality when stratified by age, sex and geographic regions which are crucial factors to understand the health inequities. Investigations are warranted to address how mortality of different causes may provide greater insights into the challenges of public health.

Guangdong is one of the leading developed provinces, with 108 million permanent residents that account for 8.03% of the population in China. Compared with other provinces, Guangdong achieved the highest gross domestic product (7281 billion) in 2015; however, considerable economic unbalance also existed in different regions within Guangdong, where there have been marked demographic and epidemiological transitions. Although the overall population health (represented by the under-five mortality rates and life expectancy at birth) has been improving rapidly, we noted an increasing burden of NCDs and ageing.6 7 Despite rapid healthcare transitions, little is known regarding the leading healthcare issues in different geographic regions of China.6 The profile of causes of mortality might even vary considerably when stratified by the age, sex and geographic region within a province where the magnitude of economic development and healthcare conditions and lifestyles differ considerably.

In this study, we sought to examine the health transition for 110 causes of death in Guangdong by stratification of age, sex and regions between 2005 and 2015.

Methods

The methods for analysing the findings have been described previously.9–11 Briefly, this study focused on 110 causes of death, including diseases and injury in Guangdong between 2005 and 2015. Our analysis was performed with the 129 county-level administrative units for all-cause and cause-specific mortality estimation. Original data sources were recalculated for county-level mortality because of the changes in administrative units. In this study, we provided a broad explanation of the mortality analysis by addressing the health transition in different sex and age subgroups.

Province-level and city-level all-cause mortality estimation

The main data source for mortality estimates of Guangdong was derived from Disease Surveillance Points, the Chinese Center for Disease Control and Prevention Cause of Death Reporting System, Chinese vital registration system and under-reporting of death survey.6 12 Death registration is a whole population and all-cause surveillance in Guangdong province, covering all the counties and cities. The method for provincial-level and city-level all-cause mortality estimation included: (1) calculation of the rate of under-reporting for more accurate estimates of mortality because of incompleteness of data in each system; (2) calculation of under-five mortality and 15–59-year adult mortality according to WHO’s life tables2; (3) derivation of all-cause mortality and 95% CI at county levels based on mixed-effect model that included the socioeconomic variables of the lagged distributed income, duration of education, urbanisation rate and under-five mortality and 15–59-year adult mortality.2 13–16 The quality of all-cause mortality adjusted by the completeness was deemed eligible when being within 95% CI. Ineligible all-cause mortality at county level would be replaced by the estimation of mixed-effect model; (4) inclusion of the qualified mortality rates at county levels to estimate for sex-specific under-five mortality and 15–59-year adult mortality for all counties based on spatiotemporal locally weighted smooth regression (S-T LOESS) model; (5) estimation of age-specific mortality for specific age-groups by modelling life tables from the summary measures of mortality on children and adults (5q0 and 59q15, respectively); these estimates would be rescaled to match for the 2015 GBD provincial-level estimates. Finally, provincial age-specific mortality at different age-group and city-level estimates were obtained from the above-mentioned rescaled county level mortality.

Cause-specific mortality estimation

We identified and redistributed the International Statistical Classification of Diseases and Related Health Problems (ICD) codes which cannot be assigned to the underlying causes of death, those that were intermediate causes of death rather than the underlying causes, or those that lacked specificity in coding. First, deaths of unknown sex were redistributed pro-rata to known sex death from cause-age groups, and the deaths of unknown age were redistributed pro‐rata to the known age of death from cause-sex groups. Deaths coded to garbage codes were reassigned using previously published methods.10 11 17 Cancers with unspecified site (C76, C80, C97) were redistributed pro-rata to mouth and oropharyngeal cancer, oesophageal cancer, gatric cancer, colorectal cancer, melanoma and skin cancer, breast cancer, cervical cancer, corpus cancer, prostate cancer, urinary bladder cancer, lymphoma, multiple myeloma and leukaemia. Ill-defined cardiovascular causes were redistributed to IHD and other cardiovascular causes of death. Ill-defined injury codes (Y10-Y34, Y87.2) were redistributed pro-rata to specific death cause of injury. Additionally, we redistributed deaths coded to symptoms, signs and ill-defined conditions pro-rata to all non-injury causes of death, and injuries with undetermined intent pro-rata to all injury causes of death. Because of potential bias, data sources with more than 5% of ICD code inaccuracy were excluded.

Crude mortality greater than 3‰ at city levels was included in cause-specific mortality estimation, or otherwise recalculated with S-T LOESS model that included the lagged distributed income, the duration of education, urbanisation rate and year as covariates.2 13–20 Details are available in online supplemental file.

Results

Overall and specific age-standardised mortality and trends

The all-cause age-standardised mortality was 727.21 per 100 000 population in 2005. Compared with 2005, the total age-standardised all-cause mortality decreased by 19.88% in 2015 (582.63 per 100 000 population). Similar trends of decline were noted for CMNN, NCDs and injuries from 2005 to 2015 (42.80 vs 36.48, 632.18 vs 513.43 and 52.23 vs 32.72 per 100 000 population, respectively). Of all-cause age-standardised mortality, 88.12% and 86.93% was attributable to NCDs in 2015 and 2005, respectively. Further details are shown in online supplemental E-Table 1.

NCDs were more common among the leading causes of death, accounting for 18 of the top 25 causes. The top 10 causes of death in 2005 followed the order of cerebrovascular disease (CVD), chronic obstructive pulmonary diseases (COPD), IHD, lung cancer, liver cancer, road injury, stomach cancer, nasopharyngeal cancer, hypertensive heart disease and Alzheimer disease. By 2015, COPD ranked third, whereas IHD ranked third in 2005 and second in 2015 (figure 1).

Figure 1

Top 25 causes of mortality for total population in Guangdong province in 2005 and 2015. n%: changes of the death number; c%: changes of the death rate; a%: changes of the age-standardised death rate. COPD, chronic obstructive pulmonary diseases.

Among men, the top 10 causes were CVD, IHD, COPD, liver cancer, lung cancer, road injury, stomach cancer, nasopharyngeal cancer, oesophageal cancer and chronic kidney disease in 2005. However, the sequence was IHD, CVD, COPD, liver cancer, lung cancer, stomach cancer, hypertensive heart disease, road injury, nasopharyngeal cancer and Alzheimer’s in 2015 (figure 2). Among women, the top 10 death causes were CVD, COPD, IHD, lung cancer, hypertensive heart disease, Alzheimer’s disease, breast cancer, liver cancer, diabetes mellitus and stomach cancer in 2005. However, the sequence was CVD, IHD, COPD, hypertensive heart disease, Alzheimer’s disease, lung cancer, breast cancer, chronic kidney disease, lower respiratory infections and stomach cancer in 2015 (figure 3).

Figure 2

Top 25 causes of mortality for men in Guangdong province in 2005 and 2015, with the median percentage change from 1990 to 2010. n%: changes of the death number; c%: changes of the death rate; a%: changes of the age-standardised death rate. COPD, chronic obstructive pulmonary diseases.

Figure 3

Top 25 causes of mortality for women in Guangdong province in 2005 and 2015, with the median percentage change from 1990 to 2010. n%: changes of the death number; c%: changes of the death rate; a%: changes of the age-standardised death rate. COPD, chronic obstructive pulmonary diseases.

Stratification of mortality by city and geographic regions

The age-standardised mortality of 21 cities in Guangdong province between 2005 and 2015 is shown in figure 4. The age-standardised mortality progressively decreased from 2005 (879.12 vs 541.06 per 100 000 population) to 2015 (733.51 vs 449.90 per 100 000 population) in 21 cities. The age-standardised annualised mortality of Pearl River Delta (Guangzhou, Dongguan, Foshan, Zhongshan, Zhuhai and Shenzhen city) was significantly lower (all p<0.05) than that in other regions, especially in west Guangdong (Zhaoqing, Maoming and Yunfu city). In 2015, the age-standardised mortality of Guangzhou, Dongguan, Foshan, Zhongshan, Zhuhai and Shenzhen city was 504.58, 480.79, 461.31, 476.91, 487.98 and 449.90 per 100 000 population, which was significantly lower compared with that in Yunfu, Zhaoqing and Maoming city (733.51, 688.08 and 698.49 per 100 000 population, all p<0.05) (figure 4 and online supplemental E-Table 2).

Figure 4

Age-standardised mortality rate in 21 cities in Guangdong province, China, 2005–2015.

Trends of mortality stratified by sex and age

Mortality for both men and women demonstrated a similar pattern of change over time. Among men, the age-standardised mortality was 921.54 per 100 000 population in 2005 and progressively reduced by 16.98% in 2015 (765.06 per 100 000 population). There was also a progressive decline in mortality annually among women (548.42 and 407.91 per 100 000 population). The mortality increased drastically with age in both men and women, particularly among population aged 65 years or greater (online supplemental E-Figure 1 and E-Table 3).

The trends of change in mortality between 2005 and 2015 varied considerably with age. All age groups showed a decreasing trend of mortality from 2005 to 2015. The 15–49-year group initially plateaued between 2005 and 2010, and subsequently declined between 2011 and 2015. For population aged 5 years and greater, the mortality increased substantially as the age increased. In 2015, children aged 5–14 years had the lowest mortality (23.56 per 100 000 population), while adults aged 70 years and greater demonstrated the highest mortality (6603.81 per 100 000 population) (online supplemental E-Figure 2 and E-Table 4).

Health transition in different age groups

We noted considerable health transitions in different age groups. The mortality of neonatal causes, diarrhoea, pneumonia, and other infectious causes and injuries in children declined in 2015, compared with 2005. Meanwhile, CVD and cancer still ranked top of all death causes at middle age and among the elderly between 2005 and 2015, with a notable peak of cancer death (47.29%) in 55–59-year group and CVD death (50.05%) in the 80+ year group in 2015. Notably, there was a modest change in the number, age pattern and composition of injury, shifting to transport injuries in younger ages and unintentional injuries (particularly, fall) in the elderly. We also noted a substantial decline in intentional injuries (particularly, self-harm and interpersonal violence) (figure 5 and online supplemental E-Table 5).

Figure 5

The trend of death (per 100 000 population) for the causes of main death in different age groups between 2005 and 2015 (A) 2005; (B) 2015

Time-course changes in of mortality

There was a trend of reduction in chronic respiratory disease (107.31 vs 61.83 per 100 000 population), cirrhosis (15.77 vs 10.12 per 100 000 population), and diabetes, urogenital, blood and endocrine disease (29.66 vs 24.57 per 100 000 population) between 2005 and 2015. Despite declining mortality of cardiovascular diseases (260.96 vs 226.56 per 100 000 population), CBD (139.89 vs 105.11 per 100 000 population), cancer (176.65 vs 148.69 per 100 000 population) and digestive diseases (11.36 vs 7.58 per 100 000 population), these diseases still dominated the top ranking of mortality between 2005 and 2015. Notably, there was a slight increase of neurological disorder-related death (21.63 vs 25.61 per 100 000 population) between 2005 and 2015 (online supplemental E-Figure 3 and E-Table 6).

Discussion

Main findings

We have demonstrated a substantial health transition between 2005 and 2015 for 110 causes in Guangdong, a developed and culturally diverse province in south China. Our key findings included: (1) the age-standardised mortality rates of all causes, NCDs, CMNN and injury declined progressively; (2) neonatal diseases, diarrhoea, pneumonia and other infectious causes and injuries declined markedly in children; (3) NCDs still accounted for a substantial healthcare burden; (4) all-cause mortality varied substantially across geographic locations, with the lowest and highest levels in Pearl River Delta and west Guangdong, respectively; (5) there was a slight increase of neurological disorders between 2005 and 2015. Additionally, we noted modest changes in the number, age pattern and composition of injury, shifting to transport injuries in younger ages and unintentional injuries among the elderly. Finally, we noted a marked decline in intentional injuries, particularly self-harm and interpersonal violence.

Interpretation of findings

Despite marked economic development, there existed significant gaps among different cities and geographic regions. Cause-of-death patterns were also highly variable. Hitherto, NCDs remain the major causes of death. In view of the demographic trends and ageing issues in Guangdong, the changes in mortality patterns is likely to persist with major healthcare implications.

In this study, we stratified Guangdong province into three regions. Cities in Pearl River Delta (ie, Guangzhou, Dongguan, Shenzhen) had lower overall mortality than other regions. This might have been attributable to the higher socioeconomic status and better healthcare services leading to improved disease control and prevention. However, the highest mortality in west Guangdong (ie, Yunfu, Maoming, Zhanjiang) might be associated with the relatively low socioeconomic status (particularly, the disadvantage of healthcare services and shortage of medical resources).

Our study has revealed some crucial health issues in different age groups in Guangdong. Despite a notable reduction in neonatal diseases, diarrhoea, pneumonia, and other infectious causes and injuries in children, the overall mortality remained steadily high for cardiovascular diseases, CVDs, cancer and neurological disorders. This finding is line with the previous study documenting major shifts of mortality from communicable disease to NCDs.21 The burden of NCDs attributable to the high-risk behaviours of diets low in fruit, high in sodium and low in whole grains, smoking, alcohol and physical inactivity is still increasing.22 23 Similar to the nationwide findings,24 25 the prevalence of smoking in Guangdong was 47.8% in men and 1.7% in women in 2013. Secondhand smoke exposure has reached to 66.8% of the population.24 Data from the national nutrition surveys showed that the dietary habit was dominated by food low in fruit, whole grains, nuts and seeds, and high in sodium and red meat in Guangdong. The rapid development and urbanisation may have predisposed to the increased healthcare burden associated with westernisation of dietary patterns in Guangdong.7 Furthermore, CVDs remain the leading cause of death in Guangdong.6 Notably, mortality of the top 25 causes decreased markedly, except for IHDs (which ranked second of the list, with only 0.27% reduction of the mortality between 2005 and 2015). Our findings call for greater vigilance to screening of the causes and implementation of effective interventions for disease control and prevention. Compared with 20.0% in 2010, the prevalence of hypertension increased to 21.4% in 2013, and the ranking of hypertensive heart diseases steadily increased, reflecting the suboptimal diagnosis, self-management, health education and promotion, and treatment in the primary healthcare centres.

Limitations and strengths

First, the estimates may still be conservative because of the incompleteness of the report. Second, the estimates for regions with extremely low mortality have been substituted by model estimation from representative regions. Third, the quality of certification and coding as assessed through the fraction of garbage codes varied substantially across cities. Fourth, there existed considerable differences for estimates of certain types of cancer from different source of data. Finally, we did not include the uncertainty interval of the mortality in this study, regarding that death registration is a whole population and all-cause surveillance in Guangdong province. However, we should be cautious about the uncertainty of model specification, model parameter estimation and garbage code redistribution algorithms in this study.

Nonetheless, our study remains robust because the data quality has been improved significantly thanks to better training of healthcare staffs. The cooperation among Guangdong Health Commission, Guangdong Bureau of Public Security and Guangdong Bureau of Civil Affairs has minimised the risks of underestimation via regular data exchange mechanism, better death report and surveillance. Furthermore, trainings of assessment of the chain of death, cause-of-death assessment and death certificate signature were held twice annually. The garbage codes have been redistributed by following the methods of WHO, in which a rigorous analytical strategy combined with the up-to-date ICD-10-coded cause of death from the national reporting system was employed. Finally, the death report cards have been randomly selected and audited from each city at each quarter, which further safeguarded the accuracy of data reporting.

Significance and future research directions

Our findings have elucidated major healthcare implications and highlighted future directions of research and practice.

First, population ageing may have predisposed to the increasing age-associated mortality that correlated with a greater prevalence of chronic diseases. The rapid rise of mortality associated with NCDs calls for greater social efforts to jointly minimise the burden through collaboration among healthcare staffs, government and the communities.

Second, despite rapid economic development, Guangdong is confronting with major issues such as inequity of distribution of healthcare facilities and other resources. The government has invested on improving community-based healthcare facilities and services such as the establishment of early detection and management of hypertension, hyperlipidaemia, and screening and management of diabetes and impaired glucose tolerance; tertiary prevention of NCD complication, tumour palliative care. However, these represented the initial steps towards refined community-based integrated management, which will require ongoing efforts and investment.

Third, our estimates have revealed important unmet needs to improve the management of major common NCDs. Our findings may justify behavioural interventions (ie, lifestyle changes) which may reduce the risks of various NCDs.7 Previous studies have suggested that modest reductions in the ‘treatable’ risk factors might result in substantial gains in public healthcare.26–29 For instance, to combat with smoking-associated NCDs, subsidies, taxation, regulation and information campaigns should be prioritised to reduce tobacco and secondhand smoke exposure. Encouragement of consumption of fruits, whole grain, nuts and seeds, along with restriction on salt intake and alcohol consumption and an increase in physical activity may have played a major role in minimising the mortality associated with many systemic NCDs.

Fourth, our results highlighted the need to minimise the mortality associated with road injury, falls, self-harms and drowning. Unfortunately, many proven interventions to prevent major causes of unintentional injuries are underrepresented in Chinese laws and regulations.30 Policies that established accountability might help reduce the non-disease-related mortality. Road injury, self-harms and drowning could be prevented by provision of expert counselling, enforcement of stringent laws regarding use of helmets for motorcyclists and bicyclists, car seats and booster seat use for child motor vehicle passengers, speed limit laws.31–35 These intervention strategies may be worthwhile for promotion in community-based healthcare settings and different governmental sectors.

Conclusion

We have documented a dramatic change in the overall mortality and age-specific, sex-specific and cause-specific mortality in Guangdong province between 2005 and 2015. Our findings highlight important unmet needs to refine healthcare services by taking into account the inequity of age, sex and geographic regions. Identification of the ‘treatable’ risk factors that may help reduce mortality associated with the adverse lifestyles. Improved disease surveillance should be enforced and the risk factors of NCDs should be continuously monitored to minimise the overall and cause-specific mortality.

Main messages

  • This the first ever systematic analysis of city level burden of disease in southern China, demonstrating the patterns of transition of health burden for 110 causes of death by stratification of age, sex and geographic regions in Guangdong between 2005 and 2015.

  • The significant burden of non-communicable diseases remains a major healthcare issue despite the notable progress in reducing the mortality, requiring an ongoing effort by the government to refine the healthcare system.

  • Consideration of city-level trends will be crucial to tackle with the diverse health challenges confronted by the local governments.

  • There were spectral transitions in different categories of disease (i.e injury, neurological diseases), calling for the refinement in the health policies for local regions.

  • In order to fully use the data, we developed methods specific to this study, including China-specific garbage code redistribution and rural and urban weighting.

Current research questions

  • What’s the major cause of death during these 10 years in southern China, especially Guangdong, one of the most leading developed provinces.

  • What’s the health transition of health burden for 110 causes of death by stratification of age, sex at city-level in Guangdong?

  • How to fully use the data specific to this study?

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Acknowledgments

We thank Professor Mai-geng Zhou with the working stuffs in National Center for Chronic and Non-Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, for their help in methodology, data analysis of this study. We thank Professor Wei-jie Guan (Guangzhou Institute of Respiratory Health, The First Affiliated Hospital of Guangzhou Medical University, Guangzhou, China) for his insightful suggestions and linguistic revision of the manuscript. We thank the research team of Global Burden of Disease Study from the University of Washington, for reference to their wonderful graphic design and colour patterns.

References

Supplementary materials

Footnotes

  • X-yZ, QY, Y-jX and X-yZ are joint first authors.

  • Contributors X-jX, Y-jX and L-fL conceived of the study and provided overall guidance. X-yZ, QY, X-yZ and L-fL prepared the first draft and finalized the manuscript based on comments from all other authors and reviewer feedback. X-yZ, QY, X-yZ, Y-jX and L-fL played a key role in formulating the analysis. All other authors contributed to the analysis and reviewed the manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.