Author + information
- Received March 10, 2016
- Revision received May 16, 2016
- Accepted June 9, 2016
- Published online August 23, 2016.
- Yanping Li, PhDa,
- Dong D. Wang, MD, DSca,
- Sylvia H. Ley, PhDa,
- Annie Green Howard, PhDb,
- Yuna He, PhDc,
- Yuan Lu, DScd,e,
- Goodarz Danaei, DSce,f and
- Frank B. Hu, MD, PhDa,f,g,∗ ()
- aDepartment of Nutrition, Harvard T.H. Chan School of Public Health, Boston, Massachusetts
- bDepartment of Biostatistics, Gillings School of Global Public Health, University of North Carolina, Chapel Hill, North Carolina
- cDepartment of Nutrition Surveillance, National Institute of Nutrition and Health, Chinese Center for Disease Control and Prevention, Beijing, China
- dYale/Yale-New Haven Hospital, Center for Outcomes Research and Evaluation (CORE), New Haven, Connecticut
- eDepartment of Global Health and Population, Harvard T.H. Chan School of Public Health, Boston, Massachusetts
- fDepartment of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts
- gChanning Division of Network Medicine, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Masssachusetts
- ↵∗Reprint requests and correspondence:
Dr. Frank B. Hu, Departments of Nutrition and Epidemiology, Harvard T.H. Chan School of Public Health, 665 Huntington Avenue, Boston, Massachusetts 02115.
Background Cardiovascular disease (CVD) is a leading cause of death in China. Evaluation of risk factors and their impacts on disease burden is important for future public health initiatives and policy making.
Objectives The study used data from a cohort of the China Health and Nutrition Survey to estimate time trends in cardiovascular risk factors from 1991 to 2011.
Methods We applied the comparative risk assessment method to estimate the number of CVD events attributable to all nonoptimal levels (e.g., theoretical-minimum-risk exposure distribution [TMRED]) of each risk factor.
Results In 2011, high blood pressure, high low-density lipoprotein cholesterol, and high blood glucose were associated with 3.1, 1.4, and 0.9 million CVD events in China, respectively. Increase in body mass index was associated with an increase in attributable CVD events, from 0.5 to 1.1 million between 1991 and 2011, whereas decreased physical activity was associated with a 0.7-million increase in attributable CVD events. In 2011, 53.4% of men used tobacco, estimated to be responsible for 30.1% of CVD burden in men. Dietary quality improved, but remained suboptimal; mean intakes were 5.4 (TMRED: 2.0) g/day for sodium, 67.7 (TMRED: 300.0) g/day for fruits, 6.2 (TMRED: 114.0) g/day for nuts, and 25.0 (TMRED: 250.0) mg/day for marine omega-3 fatty acids in 2011.
Conclusions High blood pressure remains the most important individual risk factor related to CVD burden in China. Increased body mass index and decreased physical activity were also associated with the increase in CVD burden from 1991 to 2011. High rates of tobacco use in men and unhealthy dietary factors continue to contribute to the burden of CVD in China.
In the past 2 decades, China experienced a dramatic shift in diet from traditional to Western dietary patterns (1). Decreased consumption of coarse grains and legumes were countered by increased intake of animal-source food and cooking oil (2). Rapid urbanization and industrialization led to a steep decline in physical activity levels (3). These changes have been accompanied by marked increases in serum cholesterol levels (4,5), obesity (6,7), and type 2 diabetes (8,9). Smoking prevalence in China remains high (10). Overall, cardiovascular disease (CVD) has surpassed infectious diseases to become the leading cause of death in China (11). Therefore, it is imperative to study the trends in CVD risk factors and their relationship to disease burden, to evaluate current public health policies and provide guidance for future disease prevention and health promotion.
In this study, we describe time trends in dietary and other life-style risk factors for CVD from 1991 to 2011 using data from an ongoing open cohort of the China Health and Nutrition Survey (CHNS) (12). We then apply the comparative risk assessment method (13) to estimate the number of CVD events attributable to nonoptimal levels of these risk factors.
The CHNS (12) is an ongoing prospective household-based study of multiple age groups across 9 rounds of data collection, including 4,400 households with a total of 26,000 individuals in nine provinces. The CHNS was initiated in 1989 and conducted follow-up visits in 1991, 1993, 1997, 2000, 2004, 2006, 2009, and 2011. Data are publicly available for download at the University of North Carolina at Chapel Hill Population Center project website (12).
A stratified probability sampling method was applied to the study population, as described in detail previously (12). Briefly, the CHNS used a multistage, random cluster design in 9 provinces (Liaoning, Jiangsu, Shandong, Henan, Hubei, Hunan, Guangxi, Guizhou, and Heilongjiang). Within each province, 2 cities (1 large and 1 small, usually the provincial capital and a lower-income city) and 4 counties (stratified by income, 1 high-, 1 low-, and 2 middle-income) were selected. Within cities, 2 urban and 2 suburban communities were randomly selected. Within counties, 1 community in the capital city and 3 rural villages were randomly chosen. Twenty households per community were then selected for participation. We excluded data from the 1989 wave because it enrolled only adults 20 to 45 years of age. We also excluded the data for 3 megacities from the comparison of time trend, as those data were available only in 2011. Additionally, we excluded participants who were pregnant or under 35 years of age at the time of the survey.
Comparative risk assessment method
We applied population-level comparative risk assessment method to calculate the population-attributable risk (13) (i.e., the proportion of CVD burden that would have been prevented if the distribution of specific risk factor exposure had been changed to a hypothetical alternative distribution while holding other risk factors constant). We conducted all analyses separately (20 groups in total): by sex, community-level urbanization as classified by the Chinese government (urban or rural), and age group on the basis of age at measurement (35 to 44, 45 to 54, 55 to 64, 65 to 74, and ≥75 years of age). We restricted analyses to participants ≥35 years of age because of limited data on the effects of these risk factors and fewer CVD events among younger participants.
For the comparative risk assessment analysis, we included data from different sources, including: 1) the current distribution of risk factor exposure in each wave; 2) the etiological effects of risk factor exposures on coronary heart disease (CHD), ischemic stroke, and hemorrhagic stroke, respectively; 3) an alternative theoretical-minimum-risk exposure distribution (TMRED); and 4) the total number of CVD events in the population.
Risk factor selection
We selected 17 dietary and life-style risk factors for which: 1) there was sufficient evidence for the presence and magnitude of probable causal relationships with CHD, ischemic stroke, and hemorrhagic stroke; 2) there were available intervention strategies to modify exposure of risk; and 3) data on risk factor exposure were available in CHNS without systematic bias. Factors included high systolic blood pressure (SBP), high low-density lipoprotein (LDL) cholesterol, high blood glucose, high body mass index (BMI), low physical inactivity, current tobacco smoking, and 11 dietary risk factors. Table 1 summarizes these 17 risk factors, their optimal level (TMRED), CVD outcomes, and sources of the relative risk (RR) used for estimating CVD burden. As the blood samples for LDL and glucose were collected only once in the CHNS, we included all 17 risk factors for CVD burden in 2011 but only 15 factors for time trend analysis.
Sex, age, primary occupation category, highest educational level achieved, and smoking status were self-reported in each wave. Dietary information was collected by 3-day 24-h dietary recalls in addition to using the 3-day food-weighted method to assess cooking oil and condiment consumption. Nutrient intakes were calculated using the China Food Composition Tables. Specifically, Food Composition Tables-1991 (14) was used for dietary data from the 1997 and 2000 waves and Food Composition Tables-2002/2004 (15,16) was used for dietary data from the 2004, 2006, 2009, and 2011 waves. We did not include the dietary data from 1991 or 1993 because the food codes in those datasets did not match the food codes in Food Composition Tables (matching codes were not released). We evaluated energy-adjusted dietary intakes for each dietary factor, except for polyunsaturated fatty acid (PUFA), using the residual method to 2000 kcal/day (17). PUFA intake was calculated as the percentage of total energy intake.
Following standardized procedures, trained health workers measured the weight and height of all participants using calibrated equipment (SECA 880 scales and SECA 206 wall-mounted metal tapes, SECA North America, Chino, California) (18), and BMI was calculated as weight (kg) divided by height squared (m2).
Each participant’s seated SBP was measured on the right arm using standard mercury sphygmomanometers by experienced physicians who attended a 7-day data-collection training session and passed a comprehensive reliability test (19). Three measurements were obtained with a 30-s interval between cuff inflations if the first measure was normal. Otherwise, participants were requested to rest for 10 to 30 min before a second measurement was taken (19). The mean value of 3 measurements was used.
Analysis methods for blood glucose and LDL cholesterol are described in detail elsewhere (1). In brief, overnight-fasting blood glucose and LDL cholesterol were measured with the glucose oxidase phenol 4-aminoantipyrine peroxidase (GOD PAP) method and the polyethylene glycol–modified enzyme method, respectively.
Participants self-reported physical activity on a questionnaire that solicited detailed information on occupational and domestic activities (3). Total metabolic equivalent of task (MET) h/week was calculated by multiplying the MET values (20) of activities by the time spent on the activity. We also categorized study participants into different physical activity levels according to their MET min/week (0: <600; 1: 600 to 3,999; 2: 4,000 to 7,999; 3: ≥8,000 MET min) (21).
Etiological effects of risk factors on CVD
Data on etiological effects of risk factor exposures on CVD were extracted from the most recent systematic reviews and meta-analyses (13,22–35). For each risk factor-CVD association, we derived the same RR for men and women, except where empirical evidence indicated that the RR differed by sex (13,24). We used consistent age-varying distributions of RRs for CVD across risk factors, because the associations vary by age. Current evidence suggests that the effects of these risk factors on CVD in Western and Asian populations are similar (36–38), we extracted RRs from published studies regardless of the ethnicity of study populations.
Optimal exposure distribution
We used TMRED as the alternative optimal exposure distribution to assess the proportion of CVD events associated with all nonoptimal levels of exposure (13). The TMRED for smoking was zero (e.g., no tobacco smoking). For variables where the exposure of zero is physiologically impossible, such as BMI, SBP, glucose, LDL, diet, and physical activity, the levels corresponding to the lowest CVD risk in epidemiological studies or the levels observed in low-exposure populations were used as TMREDs (Table 1).
Total CVD morbidity and mortality
CVD morbidity and mortality in China were extracted from the China Health Statistical Yearbook and the National Population Census. We extracted data on age-, sex-, and urban- or rural-specific mortality rate of 2011 for CHD and stroke from the China Health Statistical Yearbook. The mortality rate was combined with the age-, sex-, and urban- or rural-specific population data from the 2010 Population Census of the People’s Republic of China to obtain CHD and stroke mortality events for each category. We applied the age- and sex-specific ratio of ischemic to hemorrhagic stroke (39) to classify total stroke deaths attributable to each type of stroke. The age- and sex-specific case-fatality (39) and mortality events attributable to CHD, ischemic, and hemorrhagic stroke deaths were calibrated to reproduce the total morbidity and mortality events for total CVD.
Mean ± SE or percentage of each risk factor was presented by sex, age group, and urban or rural area in each wave. We applied the multivariate-adjusted general linear mixed-regression model, which allowed us to account for repeated measurements collected on the same individual over time using a random intercept for subject, to calculate covariate-adjusted mean levels of risk factors for subgroups by sex, age, and urban or rural area in each wave. To quantify time trends of the risk factors, the regression models included the year of each wave as a scored trend variable. The difference in time trend of each factor between subgroups was tested by including the interactions between subgroup variables and temporal trend in risk factor in the models: linear regression models for continuous variables and logistic regression model for smoking.
To describe the overall trends for each factor over time, we estimated the distributions of the risk factor in the joint classifications of sex, age group, and urban or rural area (20 groups in total) in each wave, then calculated the sex, age, and urban or rural standardized overall distribution of each risk factor using the 2010 Chinese Population Census data as the standard.
Future trends of 15 risk factors (except LDL and glucose, for which we do not have repeated measurements) for 2012 to 2031 were projected on the basis of data for participants with >3 repeated measurements during 1991 to 2011. We used a random effects model within each stratum of age, sex, and urban or rural. We also applied a time-series analysis with Bayesian panel value-at-risk model (40,41) to project future values for 2012 to 2031 on the basis of the average secular trend of each risk factor from 1991 (for diet from 1997 and for physical activity from 2004 ) to 2011.
Estimating CVD attributable to specific risk factors
Assuming a causal relationship between each risk factor and CVD, we calculated the population-attributable fraction (PAF) (13) to estimate the proportional reduction in CVD events that would occur if risk factor exposure had been reduced to an alternative level on the basis of the total effects of the risk factor.
For risk factors measured continuously (blood pressure, LDL, glucose, BMI, and dietary factors besides marine omega-3 fatty acids), we computed PAFs using the following equation, where x = exposure level, P(x) = actual distribution of exposure in the population, P′(x) = alternative distribution of exposure in the population, RR(x) = RR of CVD at exposure level x, and m = maximum exposure level:
The discrete version of the same estimator for PAF was applied to risks measured in categories of exposure (smoking, physical inactivity, and marine omega-3 fatty acids). The PAF related to each risk factor for each outcome (CHD, ischemic, and hemorrhagic stroke) was calculated by joint classifications of sex, age groups, and urban or rural area (20 groups in total). We calculated the number of CHD, ischemic stroke, and hemorrhagic stroke events attributable to each risk factor by multiplying its PAF by total events. The number of total CVD events was the sum of the numbers of CHD, ischemic stroke, and hemorrhagic stroke events. The PAF% for total CVD due to each risk factor was estimated by dividing the sum of events attributable to CHD, ischemic, and hemorrhagic stroke by the total number of CVD events in the population. As the CVD events attributable to individual risk factors often overlap, the total CVD events attributable to all risk factors could not simply be summed.
Time trends of risk factors
On the basis of repeated measurements of the CHNS, we observed that mean SBP significantly increased over time in the whole population (Table 2), with more pronounced trends in the younger population and rural residents (both p values for interaction <0.0001). SBP increased by 2.0 mm Hg (95% confidence interval: 1.2 to 2.9 mm Hg) among urban residents and 6.7 mm Hg (95% confidence interval: 6.1 to 7.2 mm Hg) among rural residents over the 20 years. We found a similar increasing trend in BMI over time; younger and rural residents showed greater increases (Table 2). Smoking was one of the leading lifestyle risk factors for CVD, especially among men. Despite a declining trend in smoking over time, 53.4% of men were still tobacco users in 2011 (Table 2). The average physical activity level decreased from 1991 to 2011 overall (Table 2).
We classified dietary factors into 2 groups: moderation components (lower intake is preferred), including sodium, red meat, processed meat, and sugar-sweetened beverages (SSBs); and adequacy components (higher intake is preferred), including dietary fiber, fruits, vegetables, nuts, whole grains, PUFA, and marine omega-3 fatty acids.
We observed a decreasing trend in sodium intake overall, although subgroup rates differed (p for interaction >0.1 for all) (Table 2). Consumptions of SSBs, red meat, and processed meat increased over time, but overall consumption remained relatively low. Intakes of dietary fiber, fruit, nuts, vegetables, and marine omega-3 fatty acids increased significantly over time among all groups (Table 2), but average intake remained below the TMRED levels.
Disease burden of CVD in 2011
In 2011, 6.79 million Chinese residents aged 35 years or older died—3.04 million of CVD. The estimated number of CVD events was 6.97 million (CHD 2.96, ischemic stroke 2.43, hemorrhagic stroke 1.58 million) in 2011.
We estimated that high blood pressure, the leading individual attributable factor for CVD events in China, was associated with 3.1 million CVD events in 2011 (Figure 1). The estimated PAF of high blood pressure was 43.8% for overall CVD in 2011 (Online Table 1).
In 2011, high LDL cholesterol and high blood glucose associated with 1.4 and 0.9 million CVD events, respectively (Figure 1). High LDL cholesterol was the second leading risk factor for CHD, with an estimated PAF of 37.2% in 2011. High LDL cholesterol was associated with 10.8% of ischemic stroke burden. The PAF of high blood glucose was 15.1% for CHD, 10.9% for ischemic stroke, 10.7% for hemorrhagic stroke, and 12.6% for overall CVD.
In 2011, current smoking was estimated to be associated with 1.3 million CVD events with a PAF of 30.1% among men and 7.5% of women (Online Table 1). Because harmful effects of smoking persist after smoking cessation (13), we estimated another 0.28 million CVD cases would be related to former smoking in 2011 (Online Table 2). The PAF for CVD of former and current smoking together was 36.0% among men and 9.1% of women in 2011.
High sodium intake was the leading dietary factor for CVD through its effect on increasing blood pressure. In 2011, high sodium intake was associated with 1.6 million CVD events (PAF 23.5%), overlapping at least partially with the total CVD burden associated with SBP (Figure 1).
Physical inactivity was the third leading risk factor for CHD, and estimated to be associated with 0.9 million CHD events (PAF 30.8%) in 2011, followed by insufficient marine omega-3 fatty acid intake (0.8 million CHD events with PAF of 26.6%) (Figure 1). Low consumption of nuts was associated with a PAF of 9.6% for CHD, whereas low fruit consumption was associated with PAF of 19.5% for ischemic stroke and 19.2% for hemorrhagic stroke (Online Table 1).
Time trends of estimated CVD burden
A trend for increasing SBP yielded an estimated 45,264 more CVD events from 1991 to 2011, and there will be a further increase of 393,470 CVD events during 2011 to 2031 (Central Illustration, panel A). The number of CVD events attributable to high BMI was very low in 1991, but increases in BMI were associated with 584,218 CVD additional events during the following 20 years and could potentially contribute to another 310,288 CVD events from 2011 to 2031 (Central Illustration, panel C). A declining trend in current smoking was estimated to be associated with 190,826 fewer CVD cases during 1991 to 2011 and is projected to be associated with 81,329 fewer CVD cases during 2011 to 2031 (Central Illustration, panel B). In 2011, the average physical activity level was only 192 ± 7.9 MET hours/week (mean ± SE) (Central Illustration, panel D). A decline in physical activity was estimated to relate to 748,019 CVD events during 1991 to 2011, and it will associate with a further increase of 325,448 CVD events during 2011 to 2031.
The overall sodium intake decreased but remained high when compared with the TMRED level in 2011 (mean ± SE 5.4 ± 0.2 g/day; TMRED 2.0). The decreasing trend in sodium intake may result in a large reduction in CVD burden if optimal reduction in blood pressure related to sodium intake is achieved (Figure 2A). However, we might have underestimated the sodium intake by using data collected by a food-weighted method that weighted the foods, salt and other condiments consumption at home. As shown in Online Figure 1, the energy contributions from snacks, from foods prepared or eaten out of home increased over time. Including salt and condiments intake from foods prepared away from home, the average sodium intake was 6.75 g in 2011 (Online Table 2). The CVD burden associated with high sodium intake would be 2.3 million after this adjustment.
In 2011, the average intake (mean ± SE) of SSBs, red meat, and processed meat was 3.5 ± 2.4 g/day, 79.7 ± 3.4 g/day, and 4.0 ± 0.8 g/day, respectively (Figure 2). The mean intakes were 10.8 (TMRED 30.0) g/day for fiber, 67.7 (TMRED 300.0) g/day for fruits, 6.2 (TMRED 114.0) g/day for nuts, and 25.0 (TMRED 250.0) mg/day for marine omega-3 fatty acids in 2011 (Figure 3). Insufficient intakes of marine omega-3 fatty acids, fruits, and fiber were estimated to be associated with 1.3 million, 1.0 million, and 0.6 million CVD events, respectively, in 2011. Increasing intake of marine omega-3 fatty acids over time did not produce a large decline in CVD burden, because consumption was still far below optimal levels (Figure 3).
To our knowledge, this is the first population-based study of CVD burden in China that examines a wide range of dietary, life-style, and metabolic risk factors among a large and representative Chinese population. We found that multiple modifiable risk factors, including high blood pressure, high LDL cholesterol, high blood glucose, high BMI, smoking, physical inactivity, and a diet low in fruits and marine omega-3 fatty acids but high in sodium accounted for a large number of CVD events in China. We observed modest improvements in tobacco control and multiple dietary intake components, which may have slowed the rapid increase in the burden of CVD. Nevertheless, the smoking rate and dietary quality still fell short of optimal goals, and these modest improvements could not counteract the increasing burden of CVD due to unfavorable concurrent changes in BMI and physical activity levels.
The increasing upward trend in SBP was consistent with Chinese national surveillance data showing a continuous increase in the prevalence of hypertension over the past century: The prevalence of high blood pressure was 5.1% in 1959 (43), 7.7% in 1979, 13.6% in 1991 (44), 17.7% in 2002 (45), and 33.5% in 2010 (46). The most recent prevalence figures on high blood pressure are comparable to those among U.S. adults (47). The estimated PAF of 43.8% of high blood pressure implied that about 2 in 5 CVD events in China might be prevented if SBP could be managed to the theoretical minimum level of 115 mm Hg. Although sodium-attributable CVD burden was assumed to be mediated through elevated blood pressure, we did not find a parallel between decreased sodium intake and decreased blood pressure (30). One explanation is that the potential decrease in blood pressure attributable to sodium reduction may have been counterbalanced by an increase in blood pressure from higher BMI (48) and lower physical activity levels (49). Our estimation of blood pressure–attributable CVD burden was close to global estimates; high blood pressure was estimated to be responsible for 45% to 48% of deaths from ischemic heart disease and 47% to 53% of deaths from stroke in 2010 (50).
Increased BMI and decreased physical activity contributed significantly to the increase in CVD burden from 1991 to 2011 (51). China now has more obese individuals than the United States. The worldwide ranking of the number of severely obese individuals has moved China from 60th place for men and 41st place for women in 1975 to second for both men and women in 2014 (51). Our estimated CVD burden attributable to high BMI is probably conservative; etiological effect of BMI was derived mostly from Caucasian populations, and Chinese tend to have higher CVD risk at lower BMI levels than Caucasians (52,53). Meanwhile, Chinese have become more sedentary (3). Bicycle ownership decreased from 150 to 300 bicycles per 100 families in the 1990s to only 77 bicycles per 100 households in 2011. Around 61% of rural households owned a motorcycle, and 19% of urban households owned a car in 2011 (54). At the same time, TV ownership increased from 38 sets per 1,000 persons in 1985 to 112 to 135 sets per 100 households in 2011 (54). Future unfavorable trends in BMI and physical activity level will exacerbate the increase in CVD burden, especially in combination with the increasing consumption of SSBs, red meat, and processed meat. Compared to other countries such as Germany, Finland, South Korea, and India, China has experienced a more dramatic shift in dietary patterns, from a traditional diet high in plant-based foods to a Westernized, animal-based diet (55). Reduced rates of smoking and increased consumption of dietary fiber, fruit, nuts, and marine omega-3 fatty acids may have mitigated the increase in CVD, but current smoking rates are still high, and dietary quality is far short of optimal goals. Further improvement would reduce the CVD burden and disease burden from other causes (56).
High blood LDL cholesterol was the second leading risk factor for CHD. The rapid transition to a Western dietary pattern has led to rapid increase in serum cholesterol levels (4). In 1982 to 1984, the prevalence of borderline high or high total cholesterol was 17.6% in men and 19.2% in women, which increased to 24% in men and 27.1% in women in 1992 to 1994 (57). The prevalence increased to 31.3% in men and 31.7% in women in 2007 to 2008 (5). Type 2 diabetes is a growing epidemic in China, occurring at a relatively young age and low BMI (58). Type 2 diabetes was rare in China in the 1980s, with an estimated prevalence of 0.67% (41). In subsequent national surveys conducted in 1994 (59), 2000 to 2001 (60), 2007 to 2008 (9), and 2010 to 2011 (8), the prevalence of diabetes was 2.5%, 5.5%, 9.7%, and 11.6%, respectively. Those data imply that CVD events associated with high LDL and glucose will continue to increase in the future.
Our study has a number of strengths. First, it was the first population-level analysis of the CVD burden in China that included a multitude of dietary and lifestyle factors using comparable methods to examine potentially preventable risk factors for CVD. To our knowledge, only 1 previous study has estimated future CVD burden in China, but it considered only 5 factors (blood pressure, LDL, glucose, BMI, and smoking), without considering any dietary factors (39). Other disease-burden studies in China have focused mainly on mortality or quality of life (61). Second, our estimate of each RR was on the basis of the most recent and best available evidence on risk-exposure distribution in the population. Third, the exposure distribution of all risk factors was estimated on the basis of original data from the CHNS. Previous disease-burden projects were on the basis of pooling exposure distributions from different studies. The original data-based estimation allowed us to account for potential residual confounding, although it could not be completely eliminated. Other strengths included the originality and high response rate of the CHNS as this survey was the only large-scale longitudinal study of its kind in China. Overall response rates of CHNS were around 88% at the individual level and 90% at the household level (62).
CVD is likely caused by multiple factors acting simultaneously, CVD events attributable to individual risk factors often overlap, and the total CVD events attributable to all risk factors cannot simply be summed. In addition, some factors may interact. For example, physical inactivity may increase the risk of CVD through increasing BMI and blood pressure. The total number of CVD events attributable to multiple factors that may interact with each other could not be obtained by simple addition. A potential solution is to use the joint exposure distributions of all correlated risk factors together combined with their related RRs with CVD risk to estimate the overall CVD burden. However, no solidly evidence-based RRs are presently available for joint classification, including all dietary and lifestyle risk factors. We did not consider aging or population growth in our time-trend analysis of CVD burden. We applied CVD events in 2011 to all PAFs in different waves to estimate the time trend of attributable CVD burden. This may have overestimated the CVD burden before 2001 but underestimated future CVD burden, because the Chinese population is aging. As reported in 1 previous estimate (39), CVD events will probably increase by 50% from 2011 to 2031 due to aging and population growth alone, even if all risk factors remain at year-2011 levels. Our projection of CVD burden associated with individual risk factors might be an underestimation, as our standardization and estimates were on the basis of the population proportion and number of CVD events in 2011.
Our estimates of the preventable CVD burden are important for public health and policy makers in China. Previous studies demonstrate the efficacy and effectiveness of preventable strategies in reducing the levels of CVD risk factors and total CVD burden in high-risk individuals and general population (1). The Chinese Dietary Guideline, with its visual version, Chinese Food Pagoda, had been developed to promote dietary advice to the public for many years (63,64). Most of the recommendation levels of the Chinese Dietary Guidelines are in line with the TMREDs in this analysis, such as recommendations on vegetables (400 g/day) and fruit (300 g/day) (63,64). We used a TMRED level of 2 g/day for sodium, which is slightly lower than the recommended level (2.4 g/day sodium [6 g/day salt]) in the Chinese Dietary guidelines. We intended our analysis to contribute to the scientific evidence base for population-level CVD prevention strategies; recommendations on salt intake in the Chinese Dietary Guidelines considered both scientific evidence and practical health promotion in Chinese who had a very high usual intake level of sodium (19). Health education and life-style intervention programs are effective in prevention of diabetes and CVD in China (1). Effective screening strategies are needed, as less than one-third of hypertension and diabetes patients were aware their conditions. Although promotion of smoking cessation, national bans on smoking in public and work places, and bans on tobacco advertising have halted the increasing trend of smoking prevalence (65), the percentage of male smokers is still high in China; targeted interventions are still warranted. Besides managing preventable risk factors, other strategies should also be helpful for CVD control include promoting health insurance coverage, eradicating poverty, developing environments conducive to walking and bicycling, and controlling air pollution. Prevention should be a top national policy priority for CVD control (1), which depends on fully involved stakeholders from government, health care, education, industry, urban planning, the media, the food production and service sectors, nongovernmental organizations, communities, and individuals (1).
High blood pressure remains the leading factor contributing to the increasing CVD burden in China, though increasing BMI and decreasing physical activity were also important. Decreased smoking prevalence and sodium consumption, and increased fruit, fiber, and seafood intakes contributed to slight reductions in CVD burden; however, the current levels of these factors remain below optimal levels. With rapid westernization of the Chinese diet, consumption of red meat, processed meat, and SSBs should be targets for future interventions to prevent CVD in China.
COMPETENCY IN SYSTEMS-BASED PRACTICE: High blood pressure is the most prevalent cardiovascular risk factor in China, but elevated BMI and physical inactivity also contributed to increases in CVD burden from 1991 to 2011. High rates of tobacco use in men and unhealthy dietary habits are ongoing threats to the health of the population.
TRANSLATIONAL OUTLOOK: More work is needed to develop and implement prevention strategies directed at modifiable risk factors, at both the individual and population level, to reduce the mounting burden of CVD in China.
Most of the data presented in this paper were on the basis of the China Statistic Yearbook, China Public Health Statistical Yearbook, National Population Census Datasets, and 8 waves of data from the China Health and Nutrition Survey. The authors thank Ms. Guifeng Jin and Dr. Shufa Du from the University of North Carolina at Chapel Hill and Huijun Wang from the China National Institute of Nutrition and Health for providing individual assistance in the data preparation.
For supplemental tables and a figure, please see the online version of this article.
This study was supported by the Swiss Re Foundation. The authors thank the National Bureau of Statistics of the People’s Republic of China, National Institute of Nutrition and Food Safety, China Center for Disease Control and Prevention, Carolina Population Center (5 R24 HD050924), the University of North Carolina at Chapel Hill, the National Institutes of Health (NIH) (R01-HD30880, DK056350, R24 HD050924, and R01-HD38700) and the Fogarty International Center, NIH, for financial support for the China Health and Nutrition Survey (CHNS) data collection and analysis files from 1989 to 2011 surveys, and the China-Japan Friendship Hospital, Ministry of Health, for support for CHNS 2009. All authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- body mass index
- coronary heart disease
- China Health and Nutrition Survey
- cardiovascular disease
- low-density lipoprotein
- metabolic equivalent of task
- population-attributable fraction
- polyunsaturated fatty acid
- relative risk
- systolic blood pressure
- sugar-sweetened beverages
- theoretical-minimum-risk exposure distribution
- Received March 10, 2016.
- Revision received May 16, 2016.
- Accepted June 9, 2016.
- American College of Cardiology Foundation
- Farzadfar F.,
- Finucane M.M.,
- Danaei G.,
- et al.
- Yang W.,
- Xiao J.,
- Yang Z.,
- et al.
- Zhang B.,
- Zhai F.Y.,
- Du S.F.,
- Popkin B.M.
- Wang G.
- Yang Y.,
- Wang G.,
- Pan X.
- Yang Y.,
- Wang G.,
- Pan X.
- Micha R.,
- Khatibzadeh S.,
- Shi P.,
- et al.
- Du S.,
- Batis C.,
- Wang H.,
- Zhang B.,
- Zhang J.,
- Popkin B.M.
- Sallis J.F.,
- Haskell W.L.,
- Wood P.D.,
- et al.
- Lim S.S.,
- Vos T.,
- Flaxman A.D.,
- et al.
- Bull F.C.,
- Armstrong T.P.,
- Dixon T.,
- et al.
- Ezzati M.,
- Henley S.J.,
- Thun M.J.,
- Lopez A.D.
- Hu D.,
- Huang J.,
- Wang Y.,
- Zhang D.,
- Qu Y.
- Law M.R.,
- Wald N.J.,
- Thompson S.G.
- Luo C.,
- Zhang Y.,
- Ding Y.,
- et al.
- Micha R.,
- Wallace S.K.,
- Mozaffarian D.
- Threapleton D.E.,
- Greenwood D.C.,
- Evans C.E.,
- et al.
- Lawes C.M.,
- Parag V.,
- Bennett D.A.,
- et al.
- Zhang X.,
- Patel A.,
- Horibe H.,
- et al.
- Moran A.,
- Gu D.,
- Zhao D.,
- et al.
- ↵Tsay RS. Bayesian analysis in financial econometrics: value at risk and high frequency data. Available at: http://www.istfin.eco.usi.ch/getentelist/seminar-papers-tsay.pdf. Accessed June 22, 2016.
- Wu Y.,
- Huxley R.,
- Li L.,
- et al.
- Global Burden of Metabolic Risk Factors for Chronic Diseases Collaboration
- Chen Z.,
- Yang G.,
- Offer A.,
- et al.
- ↵National Bureau of Statistics of China. China statistical yearbook. 1996–2012. Available at: http://www.stats.gov.cn/tjsj/ndsj/. Accessed June 22, 2016.
- Wang D.D.,
- Li Y.,
- Chiuve S.E.,
- Hu F.B.,
- Willett W.C.
- Wu Y.
- Pan X.R.,
- Yang W.Y.,
- Li G.W.,
- Liu J.
- Popkin B.M.,
- Du S.,
- Zhai F.,
- Zhang B.
- Xu X.,
- Hall J.,
- Byles J.,
- Shi Z.