Abstract
Objectives
This study aimed to quantify the global cardiovascular disease (CVD) burden attributable to diet low in fiber among adults aged 60 years and older using data from the Global Burden of Disease (GBD) Study 2019.
Methods
We extracted data on CVD mortality, disability-adjusted life-years (DALYs), and risk-factor exposures from the GBD 2019 study for people aged 60 and older. Age-period-cohort models were used to estimate the overall annual percentage change in mortality and DALY rate (net drift, % per year), mortality and DALY rate for each age group from 1990 to 2019 (local drift, % per year), longitudinal age-specific rate corrected for period bias (age effect), and mortality and Daly rate for each age group from 1990 to 2019 (local drift, % per year). And period/cohort relative risk (period/cohort effect).
Results
From 1990 to 2019, global age-standardized cardiovascular disease (CVD) mortality rates attributable to low dietary fiber intake decreased by 2.37% per year, while disability-adjusted life years (DALYs) fell by 2.48% annually. Decreases were observed across all sociodemographic index regions, with fastest declines in high and high-middle SDI areas. CVD mortality and DALY rates attributable to low fiber increased exponentially with age, peaking at 85–89 years, and were higher in men than women. Regarding period effects, mortality and DALY rates declined since 2000, reaching nadirs in 2015–2019. For birth cohort patterns, risks attributable to low fiber intake peaked among early 1900s births and subsequently fell, with more pronounced reductions over time in women.
Conclusions
Low dietary fiber intake is a leading contributor to the global cardiovascular disease burden, accounting for substantial mortality and disability specifically among older adults over recent decades.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-024-19897-6.
Keywords: Global burden of disease, Cardiovascular disease, Mortality, Disability-adjusted life years, Age-period-cohort models, GBD
Introduction
Cardiovascular disease (CVD) is one of the leading causes of death and disability globally [1]. According to WHO, nearly 18 million people died from CVD worldwide in 2019 [1, 2]. The incidence and mortality of CVD have been increasing globally, especially in low- and middle-income countries [3]. The main drivers are the aging population and the prevalence of unhealthy lifestyles [4, 5]. Hypertension, high cholesterol, obesity and diabetes are the major risk factors for CVD [4, 5]. While a low-fiber diet alone may not completely prevent cardiovascular disease, significantly increasing dietary fiber intake will greatly reduce the incidence of CVD, as other risk factors such as air pollution are also closely associated with cardiovascular disease [4, 5]. Current research on CVD epidemiology is focused primarily on identifying new risk factors, assessing the CVD burden across different world regions, investigating the relationship between CVD and other diseases (such as cancer), and evaluating high-risk factors for CVD in diverse populations [1, 6, 7]. Large-scale cohort study data have been analyzed in many countries and regions, providing valuable resources for CVD epidemiological research [6–8]. Concurrently, advancements in statistical and bioinformatics methods, such as multivariable adjustments and machine learning models, have further facilitated progress in CVD epidemiology.
In addition to behavioral risk factors such as insufficient physical activity, smoking and alcohol consumption, a suboptimal diet is also a major risk factor contributing to the CVD burden [9]. One dietary factor is low fiber intake. Dietary fiber is derived primarily from plant foods and is defined as complex carbohydrates that cannot be digested or absorbed by the human body [9–11]. As the seventh most important nutrient for humans, dietary fiber is an integral part of a healthy diet [9]. Adequate intake is known to help prevent and slow the progression of certain diseases like colorectal cancer [10, 11]. The “dietary fiber hypothesis” was first proposed in the 1970s [12]. Dietary fiber is considered an essential component of a healthy diet by many nutritional guidelines, with most countries currently recommending 25 to 35 g per day for adults [13]. Epidemiological evidence consistently suggests that low-fiber diets correlate with an increased risk for multiple health issues, including CVD, cancer, obesity, and type 2 diabetes [4, 5].
A deeper understanding of the burden of cardiovascular disease due to low fiber intake could help disease control authorities develop targeted prevention programs to increase high fiber intake. The 2019 Global Burden of Disease (GBD) quantifies the burden of 369 diseases and injuries and 87 risks in 204 countries and territories from 1990 to 2019. Using the 2019 Global Burden of Disease (GBD) data, age-period-cohort (APC) model was used to analyze the changing trends of CVD attributable to low fiber intake at global, regional and country levels from 1990 to 2019.
Methods
Data sources
The GBD 2019 study is a large international collaboration, supported by the Institute for Health Metrics and Evaluation and sustained by ongoing cross-country collaboration. These available epidemiological data have been used to estimate the global burden of 369 diseases and 87 risk factors by age and gender in 204 countries and regions from 1990 to 2019 [8, 9]. The birth cohort data was obtained from the GBD 2019 study. To distinguish between different cohorts, we used ten-year intervals starting from 1900 to 1909 up to 1950–1959. This approach allows for an analysis of generational shifts in CVD risk factors and nutrition. Repeated cross-sectional data on cardiovascular disease mortality, disability-adjusted life years (DALY) and 95% uncertainty intervals (UI) were collected by gender, age, region and country worldwide and in different regions for over 30 years. Specifically, the mortality rate refers to the number of deaths in the population during a specific period. DALY sums potential years of life lost (YLL) due to premature death and years lived with disability (YLD). The 95% UI reflects the certainty of an estimate, bounded by the 25th and 95th ordered values from the 1000-fold posterior distribution. We also obtained the Socio-demographic Index (SDI) for each country/region based on lagged per capita income, average years of education for those 15 + years, and total fertility rate for women under 25 years. This measure ranges from 0 to 1 and represents socioeconomic status related to health outcomes. Higher values indicate higher socioeconomic levels. Countries were divided into 5 SDI quintiles: low, moderate-low, medium, high-moderate and high.
Case definition
Exposure to a diet low in fiber was defined by the GBD 2019 as average daily consumption of less than 23.5 g per day of fiber from all sources including fruits, vegetables, grains, legumes and pulses [3, 5]. The definition of a diet low in fiber was standardized across the surveys, and the exposure level was adjusted for a 2000 kcal per day (8.37 MJ per day) diet using a residual method. International Classification of Disease-10 (ICD-10) codes were used to represent IHD (I20–I25) and CRC (C18–C21, D01.0–D01.2, and D12–D12.8) [14]. The data was collected from a variety of sources, including national health surveys, dietary intake surveys, and food balance sheets. The GBD study employs sophisticated statistical techniques to estimate the intake of dietary fiber and other nutrients, adjusting for variations in data collection methods and potential biases. The comprehensive nature of the GBD dataset allows for robust analysis of dietary fiber intake and its association with cardiovascular disease outcomes on a global scale.
Each CVD cause and related health states were identified with standard case definitions. IHD represented acute myocardial infarction, chronic stable angina, chronic IHD, and heart failure due to IHD. Myocardial infarction followed the Fourth Universal Definition and was adjusted to include out-of-hospital sudden cardiac death. Stable angina used the Rose Angina Questionnaire definition. Stroke was defined per the World Health Organization for 3 subcategories: (1) ischemic stroke (IS); (2) intracerebral hemorrhage; and (3) subarachnoid hemorrhage [15]. Lower extremity peripheral artery disease (PAD) had an ankle brachial index of < 0.9.
Statistical analysis
Overall time trend analysis of CVD mortality and DALY in the aged
The first aim of this study was to investigate the temporal trends in the mortality and DALY rate of CVD from 1990 to 2019. We used ASMR and ASDR to eliminate the effects of demographic differences. Trends in ASMR and ASDR for CVD burden due to diet low in fiber were determined by estimated annual percentage change (EAPC). The EAPC is a widely used measure of age-standardized rate (ASR) trends.To do so, we calculated ASR per 100,000 using the following formula:
 |
(a.i. : the age-specific rate in ith the age group; w: the number of people in the corresponding i.th age group among the standard population; A: the number of age groups)
To quantify such trends in a specified time interval, an ASR EAPC measurement was applied. Assuming a linear relationship between the natural logarithm of ASR and time, the regression-line-fitted rate was determined as follows:
y = α + βx + ε.
where y is ln(ASR), X is the calendar year, ε is the error term and β is the positive or negative ASR trend. The EAPC was calculated based on the formula EAPC = 100 × (exp(β) − 1)EAPC = 100 × (exp(β) − 1), and the 95% confidence interval (CI) of the EAPC value was also obtained from the linear regression model. If the EAPC estimation and the lower boundary of its 95% CI were both > 0, the ASR was considered to be growing in trend. On the contrary, if the EAPC estimation and the upper boundary of its 95% CI were both < 0, the ASR was considered to be declining in trend. Otherwise, ASR was regarded as stable.
The trend of the ASR can be discerned by analyzing the EAPC and corresponding 95% CI. An upward ASR trend is indicated if the EAPC value and lower confidence limit are positive. Conversely, a downward trend is suggested if the EAPC value and upper limit are negative. To predict disease burden from 1990 to 2045, we used a log-linear age-period-cohort model in R with the NORDPRED package. This restricts linear projection to curb exponential growth, suiting recent trends. To explore factors influencing disease burden changes, we calculated Pearson’s correlations between ASRs and SDI globally and in 20 regions from 1990 to 2019.
Meanwhile, to address the objective of understanding the burden of CVD attributable to low fiber intake across different socio-demographic contexts, we analyzed trends by region and national level. This analysis helps to identify regions and countries with the highest and lowest burdens, contributing to the objective of informing targeted prevention programs and policy interventions.
Age–period–cohort analysis
The second objective was to conduct an age-period-cohort (APC) analysis to assess the impacts of age, period, and birth cohort on CVD mortality and DALY rates. The age factor reflects social and biological dynamics of aging. Period effects refer to impacts of events and changes, such as updates in diagnostic criteria or treatment advances, on CVD statistics across age groups. Cohort effects refer to changes in disease impact due to generational differences in risk factor exposure levels. We used the APC Web Tool (National Cancer Institute; http://analysistools.nci.nih.gov/apc/) with key parameters: (a) net drift, the overall annual percentage change by calendar year and birth cohort; (b) local drift, log-linear trends for each age group by calendar year and birth cohort, detailing the annual percentage change; (c) longitudinal age curves showing adjusted age-standardized rates for the reference cohort; (d) period relative risk (RR) compared to the reference period, adjusted for age and nonlinear cohort effects; and (e) cohort RR compared to the reference cohort, adjusted for age and nonlinear period effects. Wald chi-square tests determined parameter significance. We report overall and sex- and SDI-disaggregated age, period, and cohort effects on CVD globally. Analyses and graphics used R statistical software, version 4.3.0. Two-sided p < 0.05 indicated statistical significance.
Results
Trends in CVD burden attributable to diet low in fiber
From 1990 to 2019, the global age-standardized mortality rate (ASMR) attributable to low fiber intake decreased by 2.37 per year, from 64.6 per 100 000 to 33.6 per 100 000 (Table 1; Fig. 1A and B). ASMR showed a downward trend in all SDI regions, for example, High SDI (-4.06, 95% CI: -4.22 to -3.9), High-middle SDI (-2.87, 95% CI: -3.33 to -2.4), Middle SDI (-2.06, 95% CI: -2.3 to -1.82), Low-middle SDI (-1.29,95%CI: -1.42 to -1.17), and Low SDI (-0.68,95%CI: -0.94 to -0.42) (Table 1). In contrast, the ASMR increased in Central and Central Sub-Saharan Africa (Table 1; Fig. 2A and B). At the national level, in 2019, China reported the highest CVD mortality due to low fiber intake (48132.4,95%UI: 14495.6-108858.4) and the lowest CVD mortality due to low fiber intake was reported in Tokelau, Niue, and Nauru (Fig. 2A). Meanwhile, for ASMR, Mongolia and Tajikistanreported the highest rates of 205.5/100 000 (95%CI: 111.6-311.7) and 172.4/100 000 (95%CI: 80.3-277.8), respectively. In contrast, Qatar reported the lowest ASMR (1.8 per 100 000, 95%CI: 1.2 to 4.3)(Fig. 2B). In addition, the Democratic Republic of the Congo had the largest increase in the EAPC (4.12,95%CI: 2.89 to 5.37) and Cuba, Equatorial Guinea, and Senegal had the largest decreases in the EAPC (FIG. S1A). Figure 3 A shows that ASMR is negatively correlated with SDI, although the p-value is greater than 0.05. In general, the ASMR of CVD caused by low fiber intake showed a downward trend worldwide and in all SDI regions. ASMR was consistently lower in the high SDI and high-medium SDI regions than in the global region; conversely, ASMR was higher in the low SDI, low-medium SDI, and medium SDI regions than in the global region (FIG. S2).
Table 1.
Deaths of Cardiovascular diseases between 1990 and 2019 in the old at the global and regional level
| Location | 1990 | 2019 | EAPC_95%CI | ||
|---|---|---|---|---|---|
| Number_95%UI | ASR | Number_95%UI | ASR | ||
| Global | 309319.1 (160796.7-465862.8) | 64.4 (33.5–97) | 340153.7 (179478.8-526281.3) | 33.6 (17.7–51.9) | -2.37 (-2.56–2.18) |
| High SDI | 87198.1 (42497.4-133349.9) | 65.5 (31.9-100.1) | 51809.6 (25130.5-81285.9) | 22.1 (10.7–34.7) | -4.06 (-4.22–3.9) |
| High-middle SDI | 78200.5 (36782.6-124814.2) | 58.4 (27.4–93.1) | 73776.4 (34762.3-122300.5) | 28.5 (13.4–47.2) | -2.87 (-3.33–2.4) |
| Middle SDI | 78793.9 (41784.9-120769.1) | 66.4 (35.2-101.8) | 106343.1 (55647.4-170225.8) | 35.2 (18.4–56.4) | -2.06 (-2.3–1.82) |
| Low-middle SDI | 53141.9 (31001.1-76279.8) | 78.5 (45.8-112.7) | 86539.7 (48447.1-129019) | 53.8 (30.1–80.2) | -1.29 (-1.42–1.17) |
| Low SDI | 11809.2 (5811.3-18696.4) | 44.6 (22-70.6) | 21491.6 (10545.5-33780.1) | 37.7 (18.5–59.2) | -0.68 (-0.94–0.42) |
| Andean Latin America | 1265 (618.5-1933.1) | 54.1 (26.4–82.6) | 1556.5 (757.5-2487.1) | 23.4 (11.4–37.4) | -2.8 (-3.09–2.5) |
| Australasia | 2188 (990.9–3452) | 72 (32.6-113.6) | 943.5 (421.4-1530.3) | 15.3 (6.8–24.8) | -6.17 (-6.51–5.83) |
| Caribbean | 2368 (1087.6–3717) | 74.6 (34.3-117.2) | 2043.5 (953.6-3355.8) | 32.7 (15.3–53.7) | -3.11 (-3.48–2.75) |
| Central Asia | 6801.3 (3269.6-10410.4) | 122.2 (58.7–187) | 6802.7 (2999-11226.4) | 81.4 (35.9-134.4) | -1.95 (-2.8–1.09) |
| Central Europe | 15192.3 (6931.5-24082.4) | 79.5 (36.3–126) | 12160.7 (5618.8-19824.5) | 42.7 (19.7–69.6) | -2.47 (-2.77–2.17) |
| Central Europe, Eastern Europe, and Central Asia | 49544.2 (23049.6-78591.5) | 81.1 (37.7-128.7) | 48907.2 (21883.7-80802.5) | 59.3 (26.5–98) | -1.75 (-2.42–1.08) |
| Central Latin America | 2863.2 (1413.3-4377.1) | 30 (14.8–45.9) | 5252.4 (2408.8-8839.2) | 18.6 (8.6–31.4) | -1.71 (-1.88–1.54) |
| Central Sub-Saharan Africa | 661.3 (281.4-1149.4) | 26.2 (11.1–45.5) | 2079.1 (810.6-3754.9) | 37.6 (14.7–67.9) | 1.35 (0.64–2.05) |
| East Asia | 54210.3 (19810.2-99165.6) | 51.9 (19–95) | 52059.8 (17173.4-114150.8) | 19.8 (6.5–43.4) | -2.88 (-3.24–2.52) |
| Eastern Sub-Saharan Africa | 1747.5 (811.4-2901.4) | 20.9 (9.7–34.7) | 2657.7 (1316.7-4346.8) | 15.2 (7.5–24.9) | -1.33 (-1.42–1.24) |
| High-income Asia Pacific | 8990.8 (4405.8-13994) | 35.8 (17.6–55.8) | 8908.2 (4551.5-13796.4) | 16.6 (8.5–25.6) | -2.71 (-2.87–2.54) |
| High-income North America | 33143.6 (15730.4-51450.3) | 73.5 (34.9–114) | 18497.9 (7725.9-31858.6) | 23.1 (9.6–39.7) | -4.37 (-4.54–4.2) |
| North Africa and Middle East | 8107.2 (3755.4-13207.2) | 41.3 (19.1–67.3) | 12595.6 (6031.6-20460.3) | 26 (12.4–42.2) | -1.96 (-2.16–1.76) |
| Oceania | 39.6 (19.5–65.6) | 12.1 (6-20.1) | 49.6 (28.4–89.4) | 6.8 (3.9–12.2) | -2.05 (-2.13–1.96) |
| South Asia | 47560.6 (26574.4-70350.1) | 76 (42.5-112.4) | 81727.4 (42944.8-127398.9) | 48.9 (25.7–76.3) | -1.47 (-1.67–1.28) |
| Southeast Asia | 37240.6 (23217.4-50837.2) | 129.3 (80.6-176.5) | 64468.5 (36704.3-92525.6) | 90.8 (51.7-130.4) | -1.08 (-1.24–0.92) |
| Southern Latin America | 4797 (2368-7193.2) | 82.4 (40.7-123.5) | 3543.8 (1681.3-5579.3) | 34 (16.1–53.5) | -2.79 (-2.99–2.59) |
| Southern Sub-Saharan Africa | 460 (198.6-799.1) | 14.4 (6.2–25) | 1025.3 (438.4-1868.4) | 15.6 (6.7–28.4) | 0.33 (-0.17-0.84) |
| Tropical Latin America | 7106.9 (3549.9-10684.5) | 66.7 (33.3-100.3) | 6950 (3095.9-11684.2) | 23.6 (10.5–39.7) | -3.98 (-4.2–3.75) |
| Western Europe | 44750.4 (21247.3-69920.9) | 60 (28.5–93.8) | 25066.1 (11710.7-40158) | 22.9 (10.7–36.7) | -3.69 (-3.91–3.48) |
| Western Sub-Saharan Africa | 2274.9 (1100.6-3723.8) | 22.7 (11-37.2) | 1821.7 (1008.3-2888.7) | 9.1 (5-14.4) | -3.43 (-3.65–3.21) |
Abbreviations: EAPC Estimated annual percentage change, SDI Sociodemographic Index, UI Uncertainty interval
Fig. 1.
Trends in cardiovascular disease mortality and disability-adjusted life years (DALYs) from 1990 to 2019 (A) trends in deaths; (B) Mortality trends; (C) Trends in DALY cases; (D) Trends in DALY rates
Fig. 2.
Global burden of disease of CVD attributed to low dietary fiber in 204 countries and territories (A) mortality from CVD; (B) ASMR of CVD; (C) DALY rate of CVD; (D) ASDR of CVD
In terms of DALY rate, the global age-standardized rate decreased from 1166 per 100 000 in 1990 to 595.3 per 100 000 in 2019, with an EAPC of -2.48 (95% CI: -2.67 to -2.28)(Fig. 1C and D). The countries with the largest growth trends were the Democratic Republic of the Congo (3.99%, 95%CI: 2.78–5.2) and Lebanon (2.02%, 95%CI: 1.42–2.62) and the Philippines (1.97%, 95%CI: 1.34–2.6). In contrast, the largest reduction in DALY rate due to low fiber intake was observed in Cuba(-10.98%, 95%CI: -12.28 - -9.66) and Equatorial Guinea (-10.04%, 95%CI: -11.15 - -8.91) and Senegal (− 7.87%, 95% CI, − 8.61 to − 7.14)( Fig. 2c and D, and FIG. S1B). Figure 3B shows a negative correlation between ASDR and SDI. Overall, the ASDR for CVD attributable to diet low in fiber decreased globally and across SDI regions, similar to the changes in ASMR (FIG. S2).
Interestingly, the results of gender analysis suggested that the male/female ratio of CVD due to low fiber intake decreased with age in people over 60 years old, but the male/female ratio of ASMR and ASDR increased in the 85–89 age group in the low SDI, low-middle SDI and middle SDI areas. Meanwhile, the analyses suggested that ASMR and ASDR were higher in men than in women globally and in all SDI regions (FIG. S3).
Age, period and cohort effects on the global trend
From 1990 to 2019, the rate of death and DALY from CVD attributable to low fiber intake increased with age among people aged 60 years and older (Fig. 4 and FIG. S4), and the fluctuation was more obvious among women (local drift, -2.92 among people aged 85 to 89 years; 95% CI, -3.02 to -2.83)(Table S1). Similar trends were observed among men (local drift of − 2.62 for those aged 85 to 89; 95% CI, − 2.86 to − 2.39)(Table S4). Rates of death and DALY from CVD were highest at 85 to 89 years of age for both women and men (Fig. 4; Table S1-4).
Fig. 4.
Local drift of CVD in the world and in each SDI region from 1990 to 2019. (a) Mortality; (b)DALY
The time effect showed that the mortality rate and DALY rate decreased from 1999 to 2019 in both sexes (Fig. 4). Taking the mortality rate as an example, compared with 2000–2004, the mortality rate was the lowest in 2015–2019 (0.6,95%CI: 0.59, 0.62) and the highest in 1990–1994 (1.16,95%CI: 1.14, 1.19). Similar trends were observed for DALY rates (Tables S5 and S6, and Fig. 4).
There was an overall downward trend in cohort mortality and DALY rates for successive 10-year birth cohorts from 1900 to 1909 to 1950–1959 and since the 1940–1949 cohort (Fig. 4, Table S7-8). The mortality rate and DALY risk in the early birth cohort (before 1940) were higher than those in the centralized birth cohort (1940–1959). In the cohort before the reference group (1935–1944), men had lower risks for both measures than women. However, in more recent cohorts, the risk among men was equal to or greater than that among women (Fig. 4; Tables S7-8).
Age, period and cohort effects by SDI quintiles
From 1990 to 2019, both CVD mortality and DALY rates attributable to low fiber increased with age in all SDI regions (Figure S5). These rates peaked in the 84–89 age group, with a notably more rapid increase observed after age 70. In addition, interestingly, the magnitude of change was greater in men than in women across regions, and the magnitude of change was similar in men and women (Figure S5 and Tables S9-11). In addition, the increasing trends in mortality and DALY rates with age were most pronounced in high-medium SDI regions (Figure S5).
From 1990 to 2019, the mortality rate and DALY rate of CVD attributable to low fiber showed a downward trend in all SDI areas, especially in high SDI areas and medium-high SDI areas. The rate of decline was significantly stronger in high SDI areas than in other SDI areas, and there was no significant difference between men and women. In addition, the low SDI and low-medium SDI regions showed the least change (Figure S6, Tables S12-14).
For birth cohorts after the reference group (1935–1944), changes in mortality and DALY rates were modest in the moderate-to-low SDI regions and large decreases in both in the remaining SDI regions (Figure S7, Tables S15-16).
Future burden of CVD
Figure 5 plots the predicted trajectory of CVD attributable to low fiber intake in people over 60 years of age, indicating that the global mortality and number of DALYs attributable to CVD due to low fiber intake are increasing, but ASMR and ASDR are decreasing. Notably, ASMR and ASDR in low, low-medium and medium SDI regions were higher than those in other SDI regions and the global average. The reported ASMR and ASDR in high-middle and high SDI areas were lower than the global average. In addition, there was a significant gender difference, with the trend of ASMR and ASDR in males being significantly greater than that in females (Fig. 5).
Fig. 5.
Future Forecasts of GBD in cardiovascular diseases. (A)Trend of mortality; (B) Trend of DALY rate
Discussion
Our study found that from 1990 to 2019, the global mortality and disability burden from cardiovascular disease (CVD) attributable to low dietary fiber intake decreased substantially among adults aged 60 years and older. The age-standardized mortality rate (ASMR) declined annually by 2.37%, while the age-standardized DALY rate (ASDR) fell by 2.48% per year globally. These decreasing trends were observed across all sociodemographic index (SDI) regions, with the fastest declines in high and high-middle SDI areas. In addition, we found that CVD mortality and DALY rates attributable to low fiber intake increased exponentially with age, peaking at 85–89 years. Rates were consistently higher in men than women across most age groups. Regarding period effects, we noted declining mortality and DALY rates since 2000, with the lowest risks in 2015–2019 compared to previous decades. For birth cohort patterns, CVD risks attributable to low fiber intake peaked among those born in the early 1900s and declined for younger cohorts, with more pronounced reductions over time in women.
Our findings that low dietary fiber intake is a major contributor to the global CVD burden align with prior evidence. Prospective studies and meta-analyses consistently demonstrate that higher fiber consumption lowers CVD incidence and mortality [16–18]. Each 7–10 gram per day increment in fiber intake reduces CVD events and coronary death by 10–20% [16, 17]. Additionally, dietary fiber favorably impacts other CVD risk factors, lowering blood pressure, total cholesterol, LDL cholesterol, and inflammation while improving insulin sensitivity [19–22]. Mechanistically, the viscosity and fermentability of soluble fiber types slow digestion and absorption, benefiting glycemic control [23]. The bulking properties of insoluble fibers promote satiety and healthy body weight, further decreasing CVD risks [24]. Overall, our results provide updated, robust evidence using nationally-representative data across 204 countries and 32 years that low dietary fiber intake markedly increases population-level CVD burden.
The declining temporal trends in global CVD mortality and DALY rates attributable to low fiber intake signify progress in prevention efforts. Possible reasons for these reductions include improvements in socioeconomic conditions, urbanization, education, and healthcare access facilitating better nutrition [25]. Dietary habits and fiber intake may also be improving with globalization and greater availability of fruits, vegetables and whole grains [26]. However, fiber intake remains below recommended adequate levels across most nations [27, 28]. Additionally, the rise in packaged and processed foods likely counterbalances any increase in fiber-rich staple consumption [29]. Among older adults, chewing difficulties and dental problems can also limit high-fiber food intake [30]. Therefore, while promising, further public health strategies and policies are urgently needed to promote adequate fiber consumption and curb rising CVD incidence globally [31].
This study uniquely highlights the fiber-associated CVD burden among older adults using robust age-period-cohort analytics. We demonstrate an exponential rise in low fiber-attributable CVD mortality and morbidity with advancing age, plateauing at 85–89 years. Age-related changes affecting nutrition likely underpin this pattern, including reductions in olfactory and taste functions, slowed gastrointestinal transit and absorption, and decreased intestinal microbiome diversity [32, 33]. Common age-associated conditions like dental issues, gastrointestinal disorders, mobility limitations, dementia, depression and polypharmacy may additionally hinder adequate fiber consumption [34–36]. Recent nationally-representative nutrition surveys in the United States (US), Canada and Korea similarly demonstrate substantially lower fiber intakes among older versus younger adults [37]. However, older age confers the highest cardiovascular benefits from increased dietary fiber and fiber interventions, lowering LDL cholesterol, blood pressure, arterial stiffness and inflammation more than in younger groups [38–40]. Our study therefore emphasizes the immense potential returns from improving fiber intake among older adults, who face elevated CVD risks and would gain the most from heightened consumption.
Regarding the higher fiber-associated CVD mortality and DALY rates in men, sex differences in lifestyle behaviors likely contribute [41, 42]. Evidence suggests women tend to consume more fruits and vegetables versus men across settings globally [41, 42]. Data from nutrition surveys demonstrate higher fiber intakes among US and Canadian women compared to men, which expand with age [37]. Korean women over 65 years also have greater fiber consumption than counterparts. Poorer nutrition knowledge and priorities among elderly men could underlie these patterns [43]. Additionally, the higher absolute CVD incidence in aging men magnifies any elevated behavioral risks like low fiber intake. Nonetheless, dietary improvements should target both older women and men given increasing intake remains suboptimal across groups.
The period effects demonstrate encouraging declining long-term CVD mortality and morbidity from low fiber intake over recent decades. These mirror global reductions in LDL cholesterol levels and population systolic blood pressure accompanying socioeconomic growth and healthcare improvements since 1990 [44]. However, the relative contributions of risk factor changes, treatment expansions and rising living standards warrant further investigation. Interestingly, the pace of decline was more gradual in lower SDI settings, suggesting preventive dietary changes may be slower amid resource constraints or possibly dwarfed by concomitant rises in other behavioral risks like smoking [45]. We observed that the decline in CVD mortality and DALY rates was more pronounced in high and high-middle SDI regions compared to low SDI regions.
While high SDI regions often exhibit higher CVD risks due to lifestyle factors such as increased consumption of processed foods and sedentary behavior, several factors associated with urbanization and high SDI also contribute to the decline in CVD mortality and DALY rates: (1) Healthcare Access and Quality: Urbanization often brings about improved healthcare infrastructure, leading to better prevention, early detection, and treatment of cardiovascular diseases. High SDI regions typically have more advanced healthcare systems, which can mitigate the risks associated with higher calorie and macronutrient intake through effective medical interventions and public health initiatives. (2) Education and Awareness: Higher levels of education and awareness in urbanized and high SDI regions promote healthier dietary choices, including increased fiber intake. Public health campaigns and educational programs are more prevalent in these areas, encouraging populations to adopt diets rich in fruits, vegetables, and whole grains. (3) Socioeconomic Conditions: Urban and high SDI regions often experience better socioeconomic conditions, which enable access to a diverse range of nutritious foods. This economic advantage supports the consumption of high-fiber foods, even amidst the availability of processed food options. (4) Public Health Initiatives: High SDI regions have robust public health systems capable of implementing effective dietary guidelines and interventions aimed at reducing CVD risk factors. These initiatives can significantly influence dietary patterns, leading to increased fiber consumption and reduced CVD burden.
For the cohort patterns, the elevated fiber-associated CVD burden in earlier born cohorts signifies generational shifts in CVD risk factors and nutrition. Earlier 20th century cohorts likely had lower produce availability, less nutrition knowledge and more price constraints on healthy diets than subsequent groups [46]. Public campaigns, agricultural advances, and globalized trade networks over recent decades have expanded fruit and vegetable affordability in many nations, accompanying the observed reductions in fiber-linked CVD risks [47]. Recent cohorts also benefit from childhood in increasingly obesogenic environments with more abundant, diversified food supplies and heightened awareness on dietary risks [48]. However, the high-income food environment remains unfavorable for cardiovascular health, and prevailing poor dietary habits continue to drive adverse trends in emerging nutrition-related risks like high processed food intake [49]. Sustained, lifelong efforts to improve diet quality among current and future generations are essential to curb both established and emerging behavioral determinants of rising non-communicable disease [50].
Our study has important strengths. We uniquely evaluate global fiber-associated CVD mortality and morbidity specifically among older adults, who carry the highest cardiovascular risks from suboptimal lifestyle habits but for whom data remains limited regarding behavioral disease burdens. We also leverage the extensive Global Burden of Diseases, Risk Factors and Injuries Study framework using vital statistics and health data from 204 nations over 32 years to deliver robust, reliable estimates. The employed age-period-cohort analytical approach further permits nuanced assessments disentangling temporal impacts on CVD beyond basic trends.
However, several limitations should be considered when interpreting findings. The observational study design precludes causal determinations on the relationship between low dietary fiber intake and CVD outcomes. Residual confounding from unmeasured factors influencing nutrition behaviors and cardiovascular health cannot be excluded. The dietary exposure data rely predominantly on self-reported national surveys, which can be subject to recall and social desirability biases, particularly among older respondents. Available fiber consumption information for many countries also remains sporadic over earlier decades. Long-term longitudinal cohorts with detailed, repeated individual-level dietary measurements would afford more granular insights. Additionally, the GBD database does not include detailed data on fiber intake patterns and trends, which limits our ability to set the scene fully for understanding CVD patterns. Our analyses also do not consider diet quality aspects beyond fiber intake like whole grains or refined carbohydrates that likely contribute to CVD risks. Additionally, the data sources for cohorts born in the early 1900s are relatively sparse, and secular trends over this long study period of 32 years may reflect concurrent impacts of uncaptured societal improvements expanding healthcare and infrastructure. Nonetheless, the extensive dataset across nations and decades offers uniquely broad, generalizable global evidence on dietary fiber intake as a major modifiable population-level CVD risk factor.
Conclusion
In conclusion, low dietary fiber intake is a leading contributor to the global cardiovascular disease burden, accounting for substantial mortality and disability specifically among older adults over recent decades. Encouraging decreasing temporal trends likely signify improving nutrition and living standards, but fiber consumption remains below adequate levels for most countries worldwide. Strengthening public health efforts targeting increased fruit, vegetable and whole grain intake presents major potential to achieve healthy longevity and curb rising non-communicable disease.
Fig. 3.
(A) the association between age-standardized cardiovascular disease mortality and socio-demographic indices; (B) association between age-standardized DALY rate of cardiovascular disease and socio-demographic index
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
Not applicable.
Abbreviations
- APC
Age-Period-Cohort
- ASDR
Age-standardized DALY Rate
- ASMR
Age-standardized mortality Rate
- CI
Confidence Interval
- CVD
Cardiovascular Disease
- DALYs
Disability-Adjusted Life Years
- EAPC
Estimated Annual Percentage Change
- GBD
Global Burden of Disease
Author contributions
Jia-jie Lv contributed to the conception and design of the study, acquisition of data, and initial drafting of the manuscript, coordinating the overall project. Lin-jie Zhang participated in the design and analysis of data, contributed to drafting and revising the manuscript, and assisted in data interpretation. Zhuoma Yixi was involved in data acquisition and analysis, contributed to writing the manuscript, and provided significant insights during data interpretation. Yi-chi Zhang focused on statistical analysis and interpretation of results, contributed to manuscript writing, and critically reviewed the final draft. Xin-yu Li provided expertise in data analysis and interpretation, assisted in manuscript preparation, and contributed to the discussion and conclusion sections. Cheng-hao Yang played a key role in data acquisition, study design, and participated in drafting and revising the manuscript. Ming-liang Wang supervised the project, contributed to study design and data interpretation, and critically reviewed and revised the manuscript. All authors reviewed and approved the final version of the manuscript and agree to be accountable for all aspects of the work.
Funding
There is no funding support in this study.
Data availability
Ethics approval and consent to participate: For the usage of deidentified data in GBD study, a waiver of informed consent has been approved by the University of Washington Institutional Review Board. An ethics approval and the consent to participate was not necessary.Consent for publication: All participants in this study consented to publicationAvailability of data and materials: GBD study 2019 data resources were available online from the Global Health Data Exchange (GHDx) query tool ( http://ghdx.healthdata.org/gbd-results-tool).
Declarations
Ethics approval and consent to participate
For the usage of deidentified data in GBD study, a waiver of informed consent has been approved by the University of Washington Institutional Review Board. An ethics approval and the consent to participate was not necessary.
Consent for publication
All participants in this study consented to publication.
Competing interests
The authors declare no competing interests.
Footnotes
Jia-jie Lv, Lin-jie Zhang, Zhuoma Yixi share co-first authorship.
Cheng-hao Yang, Ming-liang Wang share co-senior authorship.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Cheng-hao Yang, Email: 554616691@qq.com.
Ming-liang Wang, Email: dockj9832@163.com.
References
- 1.FISCHER MA, VONDRISKA TM. Clinical epigenomics for cardiovascular disease: Diagnostics and therapies. J Mol Cell Cardiol. 2021;154:97–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.SEBASTIAN SA. Cardiovascular disease risk communication: strategies, impact, and future directions. Curr Probl Cardiol. 2024;49(5):102490. [DOI] [PubMed] [Google Scholar]
- 3.WANG Z Q, ZHANG L, ZHENG H, et al. Burden and trend of ischemic heart disease and colorectal cancer attributable to a diet low in fiber in China, 1990–2017: findings from the global burden of Disease Study 2017 [J]. Eur J Nutr. 2021;60(7):3819–27. [DOI] [PubMed] [Google Scholar]
- 4.DUELL PB. Triglyceride-Rich lipoproteins and Atherosclerotic Cardiovascular Disease risk [J]. J Am Coll Cardiol. 2023;81(2):153–5. [DOI] [PubMed] [Google Scholar]
- 5.Global, regional, and national comparative risk assessment of 84 behavioural, environmental and occupational, and metabolic risks or clusters of risks for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017, et al. Lancet. 2018;392(10159):1923–94. J. [DOI] [PMC free article] [PubMed]
- 6.ROTH G A, MENSAH G A, JOHNSON C O, et al. Global Burden of Cardiovascular diseases and Risk factors, 1990–2019: Update from the GBD 2019 study [J]. J Am Coll Cardiol. 2020;76(25):2982–3021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Global. burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019 [J]. Lancet, 2020, 396(10258): 1223-49. [DOI] [PMC free article] [PubMed]
- 8.GAZIANO T A. Cardiovascular disease in the developing world and its cost-effective management [J]. Circulation. 2005;112(23):3547–53. [DOI] [PubMed] [Google Scholar]
- 9.ZHOU M, WANG H, ZENG X, et al. Mortality, morbidity, and risk factors in China and its provinces, 1990–2017: a systematic analysis for the global burden of Disease Study 2017 [J]. Lancet. 2019;394(10204):1145–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.PARK Y, SUBAR A F, HOLLENBECK A, et al. Dietary fiber intake and mortality in the NIH-AARP diet and health study [J]. Arch Intern Med. 2011;171(12):1061–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.KWON YJ, LEE H S, PARK G et al. Association of Dietary Fiber Intake with all-cause Mortality and Cardiovascular Disease Mortality: a 10-Year prospective cohort study [J]. Nutrients, 2022, 14(15). [DOI] [PMC free article] [PubMed]
- 12.BURKITT D P, WALKER A R, PAINTER NS. Effect of dietary fibre on stools and the transit-times, and its role in the causation of disease [J]. Lancet. 1972;2(7792):1408–12. [DOI] [PubMed] [Google Scholar]
- 13.YANG Y X, WANG X L, LEONG P M, et al. New Chinese dietary guidelines: healthy eating patterns and food-based dietary recommendations [J]. Asia Pac J Clin Nutr. 2018;27(4):908–13. [DOI] [PubMed] [Google Scholar]
- 14.Global regional. National incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the global burden of Disease Study 2017 [J]. Lancet. 2018;392(10159):1789–858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.TOLONEN H, MÄHÖNEN M, ASPLUND K et al. Do trends in population levels of blood pressure and other cardiovascular risk factors explain trends in stroke event rates? Comparisons of 15 populations in 9 countries within the WHO MONICA Stroke Project. World Health Organization Monitoring of Trends and Determinants in Cardiovascular Disease [J]. Stroke, 2002, 33(10): 2367-75. [DOI] [PubMed]
- 16.THREAPLETON D E, GREENWOOD D C, EVANS C E, et al. Dietary fibre intake and risk of cardiovascular disease: systematic review and meta-analysis [J]. BMJ. 2013;347:f6879. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.KIM Y, JE Y. Dietary fibre intake and mortality from cardiovascular disease and all cancers: a meta-analysis of prospective cohort studies [J]. Arch Cardiovasc Dis. 2016;109(1):39–54. [DOI] [PubMed] [Google Scholar]
- 18.REYNOLDS A, MANN J, CUMMINGS J, et al. Carbohydrate quality and human health: a series of systematic reviews and meta-analyses [J]. Lancet. 2019;393(10170):434–45. [DOI] [PubMed] [Google Scholar]
- 19.WANDERS A J, VAN DEN BORNE J J, DE GRAAF C, et al. Effects of dietary fibre on subjective appetite, energy intake and body weight: a systematic review of randomized controlled trials [J]. Obes Rev. 2011;12(9):724–39. [DOI] [PubMed] [Google Scholar]
- 20.REYNOLDS A N, AKERMAN A P, MANN J. Dietary fibre and whole grains in diabetes management: systematic review and meta-analyses [J]. PLoS Med. 2020;17(3):e1003053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.XIE Z, WANG L, SUN M et al. Mediation of 10-Year Cardiovascular Disease Risk between Inflammatory Diet and Handgrip Strength: base on NHANES 2011–2014 [J]. Nutrients, 2023, 15(4). [DOI] [PMC free article] [PubMed]
- 22.MAZIDI M, KENGNE A P, MIKHAILIDIS D P, et al. Dietary food patterns and glucose/insulin homeostasis: a cross-sectional study involving 24,182 adult americans [J]. Lipids Health Dis. 2017;16(1):192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.TOSH SM. Emerging science on benefits of whole grain oat and barley and their soluble dietary fibers for heart health, glycemic response, and gut microbiota [J]. Nutr Rev. 2020;78(Suppl 1):13–20. [DOI] [PubMed] [Google Scholar]
- 24.STEPHEN A M. Should we eat more fibre? [J]. J Hum Nutr. 1981;35(6):403–14. [PubMed] [Google Scholar]
- 25.HADJ AHMED S, KHARROUBI W, KAOUBAA N, et al. Correlation of trans fatty acids with the severity of coronary artery disease lesions [J]. Lipids Health Dis. 2018;17(1):52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.BELL S C, MALL M A, GUTIERREZ H, et al. The future of cystic fibrosis care: a global perspective [J]. Lancet Respir Med. 2020;8(1):65–124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.HA E J, CAINE-BISH N. Effect of nutrition intervention using a general nutrition course for promoting fruit and vegetable consumption among college students [J]. J Nutr Educ Behav. 2009;41(2):103–9. [DOI] [PubMed] [Google Scholar]
- 28.DEWEY K G, ADU-AFARWUAH S. Systematic review of the efficacy and effectiveness of complementary feeding interventions in developing countries [J]. Matern Child Nutr. 2008;4(Suppl 1Suppl 1):24–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.POTI JM. Ultra-processed food intake and obesity: what really matters for Health-Processing or Nutrient Content? [J]. Curr Obes Rep. 2017;6(4):420–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.SWEENEY J, PATTERSON C C, MENZIES-GOW A, et al. Comorbidity in severe asthma requiring systemic corticosteroid therapy: cross-sectional data from the Optimum Patient Care Research Database and the British thoracic difficult Asthma Registry [J]. Thorax. 2016;71(4):339–46. [DOI] [PubMed] [Google Scholar]
- 31.CHAMBERS T, DRAKE A, SHARPE R. Does grandparents’ diet affect weight and risk of hypogonadism in subsequent generations? [J]. Lancet. 2015;385(Suppl 1):S29. [DOI] [PubMed] [Google Scholar]
- 32.BOYCE JM, SHONE GR. Effects of ageing on smell and taste [J]. Postgrad Med J. 2006;82(966):239–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.SONNENBURG E D, SMITS S A, TIKHONOV M, et al. Diet-induced extinctions in the gut microbiota compound over generations [J]. Nature. 2016;529(7585):212–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.MOYNIHAN PJ. The relationship between nutrition and systemic and oral well-being in older people [J]. J Am Dent Assoc. 2007;138(4):493–7. [DOI] [PubMed] [Google Scholar]
- 35.WINTER JE, MACINNIS R J, WATTANAPENPAIBOON N, et al. BMI and all-cause mortality in older adults: a meta-analysis [J]. Am J Clin Nutr. 2014;99(4):875–90. [DOI] [PubMed] [Google Scholar]
- 36.PAYETTE H, COULOMBE C. Nutrition risk factors for institutionalization in a free-living functionally dependent elderly population [J]. J Clin Epidemiol. 2000;53(6):579–87. [DOI] [PubMed] [Google Scholar]
- 37.REHM C D, PEÑALVO J L, AFSHIN A, et al. Dietary intake among US adults, 1999–2012 [J]. JAMA. 2016;315(23):2542–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.KING D E, EGAN B M, WOOLSON RF, et al. Effect of a high-fiber diet vs a fiber-supplemented diet on C-reactive protein level [J]. Arch Intern Med. 2007;167(5):502–6. [DOI] [PubMed] [Google Scholar]
- 39.STREPPEL M T, ARENDS L R, VAN ‘T VEER P, et al. Dietary fiber and blood pressure: a meta-analysis of randomized placebo-controlled trials [J]. Arch Intern Med. 2005;165(2):150–6. [DOI] [PubMed] [Google Scholar]
- 40.BOULOS C, SALAMEH P, BARBERGER-GATEAU P. Malnutrition and frailty in community dwelling older adults living in a rural setting [J]. Clin Nutr. 2016;35(1):138–43. [DOI] [PubMed] [Google Scholar]
- 41.WARDLE J, HAASE A M, STEPTOE A, et al. Gender differences in food choice: the contribution of health beliefs and dieting [J]. Ann Behav Med. 2004;27(2):107–16. [DOI] [PubMed] [Google Scholar]
- 42.DENOVA-GUTIÉRREZ E, MUÑOZ-AGUIRRE P, SHIVAPPA N et al. Dietary inflammatory index and type 2 diabetes Mellitus in adults: the diabetes Mellitus Survey of Mexico City [J]. Nutrients, 2018, 10(4). [DOI] [PMC free article] [PubMed]
- 43.KOLODINSKY J, HARVEY-BERINO J R, BERLIN L, et al. Knowledge of current dietary guidelines and food choice by college students: better eaters have higher knowledge of dietary guidance [J]. J Am Diet Assoc. 2007;107(8):1409–13. [DOI] [PubMed] [Google Scholar]
- 44.Long-term. Recent trends in hypertension awareness, treatment, and control in 12 high-income countries: an analysis of 123 nationally representative surveys [J]. Lancet. 2019;394(10199):639–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.MOKDAD A H, FOROUZANFAR M H, DAOUD F, et al. Health in times of uncertainty in the eastern Mediterranean region, 1990–2013: a systematic analysis for the global burden of Disease Study 2013 [J]. Lancet Glob Health. 2016;4(10):e704–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.GONZALES-BARRON U, DIJKSHOORN R, MALONCY M, et al. Nutritional quality and staling of wheat bread partially replaced with Peruvian mesquite (Prosopis pallida) flour [J]. Food Res Int. 2020;137:109621. [DOI] [PubMed] [Google Scholar]
- 47.WEBSTER J, MOALA A, MCKENZIE B, et al. Food insecurity, COVID-19 and diets in Fiji - a cross-sectional survey of over 500 adults [J]. Global Health. 2023;19(1):99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.BHOPAL A, SHARMA S. NORHEIM O F. Balancing the health benefits and climate mortality costs of haemodialysis [J]. Future Healthc J. 2023;10(3):308–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.HALL K D, AYUKETAH A, BRYCHTA R, et al. Ultra-processed diets cause excess calorie intake and weight gain: an Inpatient Randomized Controlled Trial of Ad Libitum Food Intake [J]. Cell Metab. 2019;30(1):67–e773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.ONI T, MICKLESFIELD L K, WADENDE P, et al. Implications of COVID-19 control measures for diet and physical activity, and lessons for addressing other pandemics facing rapidly urbanising countries [J]. Glob Health Action. 2020;13(1):1810415. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Ethics approval and consent to participate: For the usage of deidentified data in GBD study, a waiver of informed consent has been approved by the University of Washington Institutional Review Board. An ethics approval and the consent to participate was not necessary.Consent for publication: All participants in this study consented to publicationAvailability of data and materials: GBD study 2019 data resources were available online from the Global Health Data Exchange (GHDx) query tool ( http://ghdx.healthdata.org/gbd-results-tool).