Article Text

Download PDFPDF

A social change perspective on injury prevention in China
  1. Joan Ozanne-Smith1,
  2. Qingfeng Li2
  1. 1 Department of Forensic Medicine, Monash University, Victorian Institute of Forensic Medicine, Melbourne, Victoria, Australia
  2. 2 Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA
  1. Correspondence to Professor Joan Ozanne-Smith, Department of Forensic Medicine, Monash University, Victorian Institute of Forensic Medicine, Melbourne, Victoria 3006, Australia; joan.ozanne-smith{at}


Introduction China has undergone massive social change over the past four decades. Since national estimates became available in 1987, the overall fatal injury rate has decreased. This paper investigates targeted interventions and sociodemographic factors that may have contributed to fatal injury rate changes particularly for road traffic fatality (RTF), suicide and drowning.

Aims (1) To review the recent epidemiology of injury in China.

(2)To investigate factors influencing trends in overall and specific cause injury mortality rates.

Methods Published injury mortality statistics and related literature were reviewed. Factors potentially influencing trends were investigated in the context of rapid development based on literature reviews of targeted interventions, macrolevel and microlevel contextual factors and changes specific to RTF, suicide and drowning.

Results Overall estimated national injury mortality rates in China decreased between 1987 and 2015, despite a rapid rise in RTF. Suicide and drowning rates decreased significantly and falls displaced drowning among the leading causes. The higher female to male suicide ratio reversed. Differences were observed in frequencies and proportions of deaths by major cause by age group and over time.

Discussion Economic and structural development and related social change in this period include: urbanisation, changes in demographics, life choices (eg, internal migration), education, poverty alleviation, health insurance and relevant regulations/legislation. These factors potentially explain much of the change in fatal injury rates in China. Data limitations persist. Increased investment in data and research would provide realistic opportunities for accelerated progress in implementing effective targeted interventions to further reduce China’s injury burden.

  • low-middle income country
  • mortality
  • economic development
  • epidemiology
  • suicide/self?harm
  • drowning

Statistics from

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.


China has undergone massive social change over recent decades.1 2 Following the Cultural Revolution (1968–1978) and the economic reforms and opening policy from the late 1970s, China experienced major and rapid changes in economics, demographics, health and politics, with profound social consequences.1 2

Between 1978 and 2006, China’s economy grew annually by 9.7% and lifted 210 million people out of poverty.3 At the same time, social welfare deficits resulted from the economic reforms and from China’s pervasive urban–rural divide and rural discontent.4–8

Over the reform era, medical care costs increased much more rapidly than incomes and close to 90% of the rural population had no healthcare insurance in 2002.9 10 Only about half of the urban population was covered, mostly formal sector workers, and not dependents or migrant workers.10

Yang et al noted improvement in disease prevention and control commenced earlier, following the formation of the People’s Republic of China in 1949.11 They describe China’s epidemiological transition as compressed into a few decades, shorter than many countries, with major causes of mortality shifting from infectious diseases and perinatal conditions to chronic diseases and injuries. They attribute this progressive change to ‘health interventions such as increased vaccination coverage; better hygiene, sanitation and water quality; improved access to medical care; and advanced social and living standards such as universal education, higher incomes, and better nutrition and housing’.11 China’s 2009 health reform plan aimed to assure that every citizen had equal access to affordable basic healthcare by 2012.10

Estimated life expectancy at birth increased from 67.9 years in 1981 to 71.4 years in 2000 and 75 years for males and 78 years for females in 2015 and infant mortality fell from 50.2 to 19.0 per 1000 live births from 1991 to 2005.3 12 The demographic shift to an ageing population was compounded by the one-child policy implemented from 1979 to 2016.13

Dramatic change at the macrolevel or structural level has been accompanied by major microlevel or individual level change. The usual public health frameworks for disease prevention applied to injury prevention, particularly the ‘public health approach’ may be inadequate to examine the effects of these changes on injury trends. While public health approaches can interpret injury mortality trends once necessary data are provided, adequate quantified data may not be available and background or contextual factors may be overlooked in applying such a disease prevention model. Sociological perspectives, accounting for both macrolevel and microlevel social change described by theorists as ‘the duality of structure’ or the micro–macro link, provide scope to consider these large-scale social processes and small-scale interactions or the role of agency.14 15 Such theories have rarely been applied to injury and its prevention.

Fatal injury rates in China were high early in the reform period,16 17 with the first national reported mortality rate for all injury of 77.0 per 100 000 persons in 1987/1988 and the three leading causes being suicide, road traffic injury and drowning.16

This paper investigates targeted interventions and other sociodemographic factors that may have contributed to changes in China’s fatal injury rates for three major causes: road traffic fatality (RTF), suicide and drowning. Within a broad macro–micro linked social change framework, it aims to (1) review the recent epidemiology of injury in China; and (2) investigate factors influencing trends in overall and specific injury mortality rates.


China does not yet have a complete death registration system, so national mortality data can only be estimated. As of 2012, vital registration data existed for only 8% of the Chinese population and Disease Surveillance Point (DSP) data for 6%.18 Since 2013 an integrated system based on vital registration and DSP covers 24% of the national population, though estimates from this system have not become publicly available.19 Methods for estimates vary from simple scale-up of the vital registration data, to complex adjustments to vital registration and DSP estimates to improve representativeness.16 18

To assess China’s injury epidemiology, published injury mortality data were obtained from multiple secondary data sources including China’s Health Statistics Year Book 2016, the WHO Global Health Estimates 2000–2015 and from the literature published during the time frame for which estimates have been available (1987–2015).17 20–26 These frequencies and rates from multiple sources were tabulated to provide a general overview. Data were included where multiple time points were available using consistent data sources and methods such as WHO Global Health Estimates and Global Burden of Disease (GBD) studies. It is because of the absence of a single report covering this entire time period and because of different methods of scaling and statistical manipulation to address data quality issues that we have included multiple estimates. Where data were relatively consistent over time, percentage changes were determined for all injury fatalities and for the major mechanisms (RTF, suicide, drowning and the increasing falls problem).17 The use of frequencies and percentages eliminates some of the additional data quality issues associated with population estimates.

The earliest published national all-cause and cause-specific injury mortality rate estimates were for 1987/1988, the first year of the vital registration system.16 18 These estimates and methodologically consistent estimates for 2005/2006 were based on vital registration data, standardised by age, sex and urban/rural location (table 1). WHO Global Health Estimates (as frequencies) were available for 2000–2015, and rates were calculated, though total death estimates may be higher based on census estimates of mortality, potentially causing underestimation in WHO mortality data, at least for 2002–2006.16 21 Comparative GBD fatal injury data were available for 1990 and 2013.17

Table 1

Selected injury mortality rate estimates per 100 000 population China (selected years and external causes)

Factors potentially influencing trends in injury mortality were investigated in the context of social change and rapid development.

Multiple literature searches were conducted on relevant sociological theory (mostly texts), contextual factors such as economic reform and social change in China including demographic, economic, educational and technological changes, as well as the agency of individual actors and groups. Searches were also conducted on national trends in injury data and trends in RTFs, suicide and drowning fatalities; relevant legislative/regulatory changes and specific targeted intervention studies for these three major causes. Reviews were limited to English language publications and English language abstracts. All searches were conducted from the period of the economic reform to the present. Articles were selected for citation where they provided new information to meet the paper’s aims and included review articles, texts and grey literature. This study did not aim to be a systematic review for several reasons including the wide-ranging topic (multiple reviews), the wide range of data sources and methods, and the potential to miss important publications if tight inclusion/exclusion criteria were applied.


The rate data shown in table 1 are from multiple selected sources, all of which have multiple time points available. All rates are estimates and no single source covered the entire period of interest. Despite the multiple sources, incompleteness of data, issues with cause of death assignment and improvements in vital registration over time, relatively consistent trends over time are seen.20

The rank order for major causes, GBD data (table 1), changed from 1990—suicide, RTF, drowning, followed by falls—to 2013 when RTF replaced suicide as the leading cause and falls replaced drowning in the top three causes (rate 8.52/100 000) and drowning moved to fourth rank. Percentage changes in rates based on GBD estimates for all injury, RTF, suicide and drowning were: −38.8,+3.5,–62.7 and −66.7, respectively.17

All-cause injury rate estimates (2000 to 2015) by sex and urban/rural residence (table 2) and external cause (table 3) were extracted from China Health Statistics Year Books by the Chinese Centre for Disease Control and Prevention (China CDC). Overall, rural rates decreased during this period, especially for women, while male urban rates increased. The injury mortality rates for 2005 suggest some artifactuality (tables 1 and 2).

Table 2

Injury mortality estimated rates (per 100 000) China, selected years, by sex and urban/rural residence

Table 3

Estimated mortality rate (per 100 000) by major injury cause: China, 2015 by sex and urban/rural residence

The RTF rate peaked for urban residents around 2005/2006, and around 2010 for rural residents, though these figures may be confounded by urban-dwelling migrant workers being recorded by their rural address and possible other issues with 2005 data.22 Economically developed, and some underdeveloped, provinces had very high RTF rates in 2005, experiencing the greatest increases between 1985 and 2005.23

Rural men had the highest mortality rates for all injury causes in all years, except 2005 when the urban male rate was slightly higher (table 2) and for all specific major injury causes in 2015 (table 3). This includes higher male suicide rates from 2006, reversing the previously higher female rates.27 28 Based on national estimates developed by China CDC from vital registration data, the rank order of major causes in 2015 was motor vehicle crashes, falls, suicide and drowning, followed by poisoning for both rural and urban place of residence and for both sexes, except for falls in urban women (ranked first) and poisoning in urban men ranked fourth (above drowning).

A 2002–2007 RTF mortality study found that the death registration rate was about double the police-reported rate.24 Linear regression showed a significant decrease of 27% in the death rate according to police sources but no significant change according to death registration data. A later study, using 5data sources, showed ongoing major discrepancies in RTF rates and trends and the results of a study investigating the likely accuracy of police versus death registration data, which simulated trends comparing road safety management performance between China and 13 other countries, suggested that the vital registration data in China may be more valid than the police data.24 29

A study of age standardised suicide trends from 1990 to 2013 reported continuous declining rates for men from 22.6 to 10.6 and for women from 24.8 to 7.7 per 100 000 per year.25 Despite this trend, the suicide rate for rural women in the older age groups did not decrease significantly from 2002 to 2011.28

A detailed study of a representative sample of 519 suicides from 23 sites in China (20 rural/3 urban) in 1998–2000 found that 62% resulted from self-poisoning with agricultural chemicals or rat poison.30 At these sites, a large proportion of rural female suicides involved: pesticide ingestion (79%), low prevalence of mental illness (39%) and a high level of acute stress from precipitatory life events, including marital infidelity or family conflict.27 28

The age distribution of injury deaths proportional to all deaths was relatively stable over the reported years 2000, 2005, 2010 and 2015 except for the age group 0–4 years, where the injury proportion increased dramatically from 12.8% in 2000 to 23.7% in 2015, while the frequency decreased (table 4).

Table 4

Injury deaths as proportion of all death estimates by age group: China selected years 2000–2015

The frequency and proportion of deaths by major cause showed some differences by age and year (table 5). The major changes over time were a reduction in drowning frequency among 5–14 and 15–29 year olds, with a little change in proportion compared with all injury deaths; a reduction in suicide frequency among 15–29 year olds; an increase in RTF and suicide frequency in 60–69 years olds (associated with a large population growth) and greatest proportion of suicides compared with all injury in this age group, but little change in percentages overall; and an increase in frequency and percentage of fall deaths in the 30–49 and 50–59 age groups (table 5).

Table 5

Major causes by age group as proportion of all injury death estimates: selected years 2000–2015


Despite limited mortality data, it is evident that China’s overall fatal injury rate has decreased during this reform period. Sufficient rigorous studies using combinations of data sources provide acceptable evidence of this downward trend. This trend is despite rapid motorisation and many other new hazards introduced by economic development such as work deaths in the construction industry and new technologies such as electric bicycle rider fatalities.31 32 Relatively few studies have provided evidence of targeted interventions contributing to these reductions.

During a period of rapid and complex change, including improved access to healthcare, epidemiological transition, reduced poverty, higher employment, increased life-expectancy and increased social welfare, injury has not declined as a proportion of all deaths in China in any age group, but has significantly increased proportionately in the 0–4 year age group and slightly in some other age groups.

The many social and developmental changes had different impacts on fatal injury rates and patterns for the specific major causes, with complex patterns of change including initial and possibly ongoing RTF rate increases and substantial decreases in suicide and drowning rates over the study period.

As in other countries, rapid motorisation in China led to a dramatic increase in RTFs.33 34 The shift from a centrally planned economy to a socialist market economy enabled rapid motorisation contributing to economic growth by improving market access and to social welfare by increasing flexibility and mobility.1

Motorisation proceeded rapidly in China from the 1990s with the main means of transportation changing from animal carts and bicycles to motor vehicles.35 Four-wheel motor vehicles increased from 60 000 in 1951 to almost a million in 1975 and 10 million in 1987. It then rose to 50 million in 1999, with an additional 30 million motorcycles.36

By 2012, China was the largest producer of vehicles and had the largest vehicle market growth world-wide.37 Ownership of motorised vehicles almost quadrupled in 10 years, from 42.2 million in 1997 to 159.7 million in 2007.38 These developments have resulted in China’s motor vehicle fleet being relatively new and potentially equipped with modern safety features.

An estimated 120 million electric bikes (E-bikes) were on the road in China in 2011, having become a dominant transportation means, with further increase expected.39 While many E-bikes resemble small motorcycles and often travel in motor-vehicle lanes, they are not regulated as motor vehicles. Increases in E-biker fatalities as high as fivefold were reported for the period 2004–2008.40

Described as of strategic importance to economic development, China completed nearly 45 000 km of high-grade toll expressways from 1990 to 2006 (from none in 1985) coinciding with rapid motorisation. Development of other road classes also continued during this period.41 Pedestrian safety measures including median barriers and overpasses and underpasses also proliferated.

In high-income countries, RTF rates peaked about 20 years after rapid motorisation commenced, post-World War II.33 Major reductions followed in high-income countries (HIC) as targeted road safety interventions such as seat belt use and traffic law-enforcement were progressively implemented.33 While there may be a recent downturn in RTF in China, the markedly different data from police and death registrations make confirmation difficult.

Although national seat-belt wearing legislation became effective in China in 2004 and associated provincial and city regulations followed, low wearing rates have been consistently reported.42–44 Since 2004, the Chinese government has adopted increasingly stricter drunk-driving policies and laws, with criminalisation from May 2011.45

Few evaluations of road safety interventions have been reported. A time series evaluation of the 2011 drunk-driving legislation (2009–2012) in Quangzhou found a corresponding significant reduction in road traffic injuries.45

Another study evaluated a 5-year drink-driving and speeding intervention in two Chinese cities (Suzhou and Dalian) from 2010 to 2014.46 The targeted interventions included social media campaigns, advocacy for legislative change and law enforcement training. Statistically significant reductions were found for both drink-driving and speeding in both cities. Although, there were no control cities and these interventions coincided with the 2011 drunk-driving legislation, the results suggest beneficial effects of the co-ordinated interventions.46

A 2010 study, citing Traffic Administration Bureau data, reported that widely installed speed radar and cameras in most cities and on express highways had proved to be effective in reducing speeding, evidenced by increases in law enforcement incidents for speeding from 12.8 million incidents in 2004 to 15.3 million in 2005 and total fines for speeding from $75.6 million (averaging $5.9 per incident) in 2004 to $155.2 million (averaging $10.1 per incident) in 2005.38

Favourable to road safety are new road networks, pedestrian treatments (barriers and over/underpasses), traffic management systems and the relatively new vehicle fleet, theoretically including modern safety features, though the extent of possible descheduling of safety features required by other markets is unknown.

While the history of legislation and enforcement for road safety is poorly documented, this has been slow, not keeping pace with motorisation and the lack of ‘proper enforcement’ has been long-recognised.47 Gaps persist in China’s road safety laws compared with world best practice.48 For example, motorcycle helmet laws are not comprehensive and there is no national child restraint law. A gap in the implementation of associated best practice interventions follows the legislative limitations.49

Dramatic reductions in China’s suicide rate with improved economic and social welfare conditions followed the high rates associated with economic reform and related social change and increase in social stressors in the 1980s.

After the Chinese government first released national suicide data in 1989, very high rates were found for the period 1990–1994 (28.72/100 000) particularly for young rural women (30% of rural female deaths aged 15–29 years) and the elderly, especially men.9 Although there are no earlier data for comparison, it has been postulated that high rates were due to social stresses occurring as the economic reforms took effect, including: economic losses for individuals and families; increasing rates of marital infidelity and divorce; rapidly increasing healthcare costs; weakening of family ties; large rural migration to urban areas for work; dissatisfaction at the increasing economic and social gap between the rich and poor and some of these changes accelerated during the 1990s.9

Different characteristics of suicide rates in China (very high rates and higher rates in females than males) compared with other countries in the 1990-1994 period may indicate unique characteristics of the Chinese environment. The rural suicide rate in this period revealed a three-fold difference between rural and urban populations.9 In addition to social factors, convenient access to lethal means of suicide was clearly important. In a largely rural sample of suicides, 62% resulted from agricultural chemicals or rat poison in 1998–2000.29

The Chinese Government ordered production of some highly lethal pesticides to stop in 2002.50 In 2005, the WHO launched a global initiative to address pesticide suicides.51 By 2008, China, Japan and Sri Lanka had introduced legislation to regulate the production, importation, transportation, storage and sale of pesticides.52 However, some of these targeted interventions appear to have been introduced after the dramatic decrease in the overall suicide rate from 1987/1988 to 2005/2006 (table 1).

The proportion of all suicides due to pesticide poisoning continued to decline from 2006 to 2013 from 55% to 49%.53 Following the progressive control of access to pesticides, improvements in medical services and transport may also have contributed to these reductions in recent years.

By 2002–2006, China’s suicide rate was significantly lower,16 particularly among young rural women and the elderly, and based on 2013 and 2015 estimates,21 26 the downward trend continued, except for rural women in the older age groups.27 The availability of greater access to affordable healthcare may have contributed to the overall suicide reduction among the elderly with healthcare costs no longer burdening the family to the extent following the initial period of economic reform. However, other social changes may have become risk factors particularly for rural groups such as social isolation, insufficient family support, loss of spouse, or debilitating or terminal illness.27

It is argued by Phillips et al that changes in suicide rates are capable of being explained by ‘the historical and dominant view in China of suicide as a social problem, related to social stressors and not primarily a mental health problem’.9

Increased life choices (agency) for younger rural women escaping extreme poverty or unbearable family environments may have affected the suicide rate, with many women gaining greater life experience, independent income and potentially less harsh conditions as migrant workers, with less direct child care responsibility due to grandparents’ increased child-rearing role, and fewer children under the one-child policy (though less restrictive in rural populations). This policy was applied most strictly to urban residents and government employees. A second child was generally allowed in rural areas and a third in some ethnic minorities and remote underdeveloped areas.13 While selective abortion or neglect and abandonment of female foetuses or infants appears to have occurred in China resulting in excess young men,54 the one-child policy effects may also have enhanced the perceived value of girls and women in the family and society.55

Improvements in healthcare, enhanced education levels, increasing public awareness of suicide and specifically urbanisation and effective control of pesticides and rodenticides13 likely contributed to the overall suicide rate reduction.

Some of the same changes, particularly economic reform, urbanisation, universal education and the one-child policy (affording responsible child care) may have influenced reductions in the drowning rate since child drowning is strongly associated with poverty, particularly with regard to parents’ educational level and number of children in the family.56

The reduction in drowning frequency, particularly in the 5–14 age group (table 5), and rate reductions (table 1) are difficult to explain by any targeted intervention since little evidence of interventions is available. A 2015 review of legislative coverage for child injury prevention in China reported that no law covered interventions to prevent child drowning deaths. There is no national programme to teach children to swim.57 Anecdotally few Chinese people can swim, though no published reports of swimming ability were identified. Hyder et al reported: a dearth of data on incidence rates and risk factors; absence of public health interventions; lack of research on intervention effectiveness and cost-effectiveness; and paucity of national drowning prevention programmes in low-income and middle-income countries including China.58

Much of the fatal drowning rate reduction may be due to economic and structural development. Urbanisation and the provision of reticulated water reduce exposure to the risk of drowning both during daily activities, such as collecting household water, bathing and washing clothes, and where child supervision is poor. Indicative China National Health Services Survey data for 1993 and 2003 show consistently high urban household access to safe drinking water (97.15% of households in 1993; 99.19% in 2003) and rural access improving from 52.49% in 1993 to 80.24% in 2003.3 Reduced family size due to the implementation of the one-child policy may have resulted in better child supervision by parents and caregivers.

A historical study of fatal drowning reductions in Victoria Australia over more than a century found that development including urbanisation and built-environment changes encompassing reticulated water, bridge construction and filling-in quarries and water holes were associated with dramatic reductions in the drowning rate. It also found that these changes preceded targeted drowning interventions.59

There are multiple limitations to this study. Mortality data for China are limited by incomplete population coverage and potentially non-representative sampling estimates both for vital registration and data surveillance points, under-reporting and misclassification, uncertain population data, including urban/rural population distribution and representation (eg, migrant workers), different methods used for scaling data to national estimates and changes over time, including improvements in coverage and quality.

Additional limitations apply to injury data by specific cause, including large discrepancies between vital registration and police estimates for RTF. There is no specific category for deaths involving electric bicycles despite their increasing numbers and crash risk. Apart from isolated studies, there is little information available on where and how drowning occurs in China, including body of water and activity, such as recreation, domestic, occupational or water transport.

It is possible that the search methods employed overlooked large scale targeted interventions or effective legislative changes which could explain reductions or stabilisation of injury rates or patterns.

Many of the large-scale or macro reforms and social changes in China have increased opportunities for individual agency or microlevel social change. Life choices have become possible due to decreased poverty, smaller families, health insurance, social mobility, migrant worker opportunities, affordable transport and mobility choices. In terms of the duality of structure, our findings suggest an increase in the roles of agency and microlevel social change over the study period. Nevertheless, macrolevel development and social change potentially explain much of China’s improvement in overall fatal injury rates, including drowning and suicide reductions, and RTF and fatal fall increases. Few large-scale targeted interventions were found to explain rate reductions and changing patterns. As supported by other authors, injury prevention efforts in China are fragmented and lack adequate research funding and many effective interventions are yet to be implemented.60 61

While data limitations persist, potentially retarding progress, sufficient information now likely exists, or could be developed with greater specificity in existing data, to inform research and further targeted interventions, based on proven countermeasures. To this end, China should invest in enhanced data and research to fill information gaps and implementation and evaluation research to drive further injury reductions. External drivers such as the Sustainable Development Goals may support action.62 These conditions now provide realistic opportunities for accelerated progress in reducing China’s injury burden.

What is already known on this subject

  • Injury rates in China have changed dramatically over the past 30 years.

  • Major economic and social reforms have taken place since 1978.

What this study adds

  • Much of the change in trends in fatal injury rates in China for the major causes: road traffic, suicide and drowning may be explained by the post-1978 reforms or consequences of the reforms.

  • Few changes in these trends can be explained by targeted interventions.

  • Injury data and research limitations continue and may be a barrier to the implementation of effective targeted interventions to reduce fatal injury rates.



  • Contributors JOS conceived of the study design, wrote the first draft and prepared the final version of the paper. QL contributed to acquisition and interpretation of data, provided critical input to drafting of the paper and approved the final version for publication.

  • Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Commissioned; externally peer reviewed.

  • Data sharing statement No additional data. New analyses of data in this paper could be conducted from published data by any other researcher.