Global disease burden and its attributable risk factors of peripheral arterial disease

Peripheral arterial disease (PAD) is a prevalent subtype of atherosclerotic cardiovascular diseases. It is crucial to assess the PAD-related burden and its attributable risk factors. We use the Global Burden of Disease study 2019 database to calculate the incidence, prevalence, mortality, disability-adjusted life years (DALY), attributable risk factors and estimated annual percentage change. The disease burden of PAD grows significantly with age accompanied by prominent heterogeneity between male and female. Despite the increase in the absolute numbers of disease burden from 1990 to 2019, the global PAD-related age-standardized death rate (ASDR) and age-standardized disability-adjusted life years rate (ASDALYR) have a mild downward trend from 1990 to 2019, which negatively correlated with sociodemographic index (SDI). Smoking and high systolic blood pressure (SBP) were the primary attributable risk factors for males (ASDR: 33.4%; ASDALYR: 43.4%) and females (ASDR: 25.3%; ASDALYR: 27.6%), respectively. High fasting plasma glucose (FPG) had become the second risk factor for ASDR (males: 28.5%; females: 25.2%) and ASDALYR (males: 29.3%; females: 26.3%) with an upward tendency. Low-middle SDI regions were predicted to have the most remarkable upward trend of PAD-related burden caused by high FPG. Smoking caused more disease burden in males before 85–90 years old and females before 65–70 years old, while high FPG and high SBP caused more burden after that. The patterns of PAD-related burden and its attributable risk factors are heterogeneous across ages, genders, and SDI regions. To reduce disease burden, tailored strategies should be implemented.


Statistical analysis
Age-standardized death rate (ASDR), age-standardized disability-adjusted life years rate (ASDALYR), agestandardized years lived with disability rate (ASYLDR), age-standardized years of life lost rate (ASYLLR), agestandardized incidence rate (ASIR) and age-standardized prevalence rate (ASPR) were showed by per 100,000 population categorized by age, gender, geographical location and social development.From 1990 to 2019, percentage changes in cases or rates were defined as follows: To assess and predict the temporal trend of age-standardized rate (ASR), the estimated annual percent change (EAPC) and 95% confidence intervals (CI) were calculated using a Generalized Linear Model (GLM) for ASR during 30 years 15 .The formula of EAPC was showed followed: where β was the annual change per 100,000 in the ASR.Positive EAPC was considered to increase ASR, while negative EAPC decreased ASR 16,17 .Regression Splines and a smoothed line were used to show the relationship between SDI and EACP in ggplot2 and R with "stat_smooth".To display the proportion of PAD cases that could be prevented if a risk was removed, population attributable fraction (PAF) was calculated.The PAF calculation method was published.The formula for PAF was shown: where A and O refer to attributable risk the observed number of cases and the expected number of cases under no exposure, respectively.
Generally, ASIR and ASPR in both genders increased with age.Both ASIR and ASPR in females overwhelmed males.However, there was a moderate decline of ASIR in females over 75-79 y/o whereas ASIR gradually increased in males (Fig. 2a-b).Both genders shared the comparable trend of ASDR and ASDALY, increasing with age (Fig. 2c-d).Generally, males had more ASDR and ASDALYR than females in all age group.Females had higher ASYLDR in all age group compared to males while opposite situation was observed in ASYLLR (Fig. 2e-f).

Trends of PAD-related burden
We calculated EAPCs separately by gender to assess the trends of PAD-related burden (ASDR, ASDALYR, ASPR and ASIR) (Fig. 3 and Supplementary Table 4).

The correlation between EAPC and SDI
Overall, the EAPC of ASIR (R-sq = − 0.57; P < 0.001), ASPR (R-sq = − 0.55; P < 0.001) and ASDALYR (R-sq = − 0.23; P < 0.001) were negatively correlated with SDI level.The EAPCs of ASIR, ASPR and ASDALYR were higher in low-, low-middle and middle SDI regions, but dropped dramatically in high-middle and high SDI regions.It indicated that PAD-related burden in lower SDI regions could be substantially underestimated (Fig. 4).

Contribution of attributable risk factors for PAD-related burden
To display PAD attributable risk factors and predict temporal trends, PAF of ASDR and ASDALYR was stratified by gender, SDI, and WBI regions to analyze individually (Supplementary Table 5).
The PAF of attributable risk factors varied significantly between male and female.In global level, smoking was the primary attributable risk factor for PAD-related burden in males (ASDR: 33.4%; ASDALYR: 43.4%; ASYLLR: 43%; ASYLDR: 44.9%) and high SBP was the primary attributable risk factor for females (ASDR: 25.3%; ASDALYR: 27.6%; ASYLLR: 26.5%; ASYLDR: 28.9%).As for ASDR, in 2019, the highest PAFs of smoking in males were in high-middle SDI regions (42.4%), upper-middle WBI regions (42.5%),Eastern Europe (54.3%),East Asia (49.6%), and Southeast Asia (46.1%).In 2019, the highest PAFs of high SBP for females were observed in high-middle SDI regions (26.4%), upper-middle WBI regions (27.3%),Western Sub-Saharan Africa (31.7%),Southern Sub-Saharan Africa (31.4%), and Southeast Asia (29.2%).The PAFs for these attributable risk factors for ASDALYR, ASYLLR, and ASYLDR in 2019 were also showed in Supplementary Table 5.The percentage change of these attributable risk factors to PAD-related burden from 1990 to 2019 were presented (Fig. 5 and Supplementary Table 5).As for ASDR in global level, smoking in males decreased by 23.49%, whereas high SBP in females decreased by 13.11%.Global ASDALYR exhibited similar tendencies.However, high FPG has become ranked 2 for ASDR, ASYLLR, and ASYLDR in males, and also grew dramatically to rank 2 for ASDR, ASDALYR, ASYLLR, and ASYLDR in females (Fig. 5).Notably, in high SDI regions, high FPG has become ranked 1 for ASDR in males whereas smoking remained to be ranked 1 for ASDALYR, ASYLLR, and ASYLDR.Furthermore, high FPG has been recognized as ranked 1 for ASDR, ASDALYR, ASYLLR in females (Supplementary Fig. 1a).However, in low SDI regions, high SBP seemed to be the predominant attributable risk factor for PAD-related burden in male and female, instead of smoking (Supplementary Fig. 1E).Regarding, the www.nature.com/scientificreports/rank for PAFs of attributable risk factors to PAD-related burden from 1990 to 2019 in different SDI quintiles were detailed in Supplementary Fig. 1.Smoking, high SBP and high FPG were the primary risk factors for PAD-related burden.To clarify PADrelated burden caused by those attributable risk factors, we classified these risk factors according to different age groups and genders.As for males, smoking contributed to the most ASDR before 85-90 y/o, while high FPG contributed to the most ASDR after 85-90 y/o.In terms of females, smoking accounted for the most ASDR before 65-70 y/o, while high FPG and high SBP accounted for the most ASDR after 65-70 y/o (Fig. 6).

Trend of attributable risks for PAD-related burden
EAPC was calculated to forecast the trends of these attributable risk factors related PAFs of ASDR and ASDALYR stratified by sex, SDI quintile and regions (Supplementary Table 5).Globally, the ASDR caused by smoking, high SBP and diet high in sodium showed downward trends  5).

Discussion
The PAD-related disease burden has been underestimated over the past three decades and significant heterogeneity was found in terms of gender, age, socioeconomic and risk factor exposure.Although ASDR and ASDALYR showed a slight downward global trend, lower SDI regions are undergoing a rapid growth of PAD-related burden compared to higher SDI regions.Gender differences were observed in PAD-related burden and attributable risk factors.High SBP was the predominant risk factor for PAD-related ASDR and ASDALYR for females, whereas smoking was the primary risk factor for PAD-related ASDR and ASDALYR for males.High FPG has become the global notable risk factor with a remarkable upward trend for PAD-related ASDR and ASDALYR.Smoking caused more disease burden in male before 85-90 y/o and female before 65-70 y/o, while high FPG and high SBP caused more burden after that.Our results might serve as an imperative extension to the present studies and provide epidemiological data and potential tailored approaches for health care policymakers and social economists.
Similar to earlier global epidemiological findings of PAD based on GBD 2017 study, our study demonstrated an increase in incidence, prevalence, mortality, and DALY cases.However, the ASRs for these four metrics decreased during the past decades.Notably, a mild decrease in PAD-related burden was observed in ASDR compared to other ASCVDs, such as stroke with 34% reduction of ASDR 18 , and ischemic heart disease with 30.8% reduction of ASDR 19 .These results indicated that PAD-related death burden had been underestimated with poor management.
The global ASIR, ASPR, ASDR, and ASDALYR of PAD showed a declining trend.Our results indicated that the PAD-related disease burden decreased as economy developed 20,21 .Moreover, the EAPCs of these four metrics were negatively correlated with SDI levels, which were consistent with the previous studies [22][23][24] .The burden of evidence was suggestive that SDI has a significant impact on the burden of cardiovascular disease through non-health determinants of health 25,26 .Our findings suggested that people in lower SDI regions may be more vulnerable to PAD, which may be a result of less uptake of fruits and micronutrients 27 .According to current studies, deficiencies in these substances, such as fiber in grains and vitamins C, D, and E in fruits and vegetables, may increase the risk of cardiovascular disease including PAD 28,29 .Specific diets have shown positive effects in preventing PAD, such as the Mediterranean diet 30 .Social determinants of health are also crucial to PAD-related www.nature.com/scientificreports/disease burden.Lower SDI regions have lower income, lower education levels, and less social support-each of which is associated with higher rates of PAD-related disease burden 31,32 .Additionally, the population in lower SDI regions has been exposed to more cardiovascular disease risks, such as smoking 33 , hypertension 34 and diabetes 35 .Prevention, detection, and treatment of PAD, such as early screening and management of major risk factors will reduce PAD-related burden 36,37 .More tailored public health strategies should be adopted in low SDI regions, such as low-cost screening for PAD, management of risk factors and standardized treatment 38,39 .We quantified the attributable burden and trends for PAD risk factors in the GBD database.In 2019, smoking, high SBP, and high FPG were the top 3 attributable risk factors for PAD-related ASDR and ASDALYR.
Diabetes was an independent risk factor for ASCVD, such as stroke, coronary heart disease, and PAD 40 .PAD patients with diabetes had a fivefold higher risk of amputation [41][42][43] .Our study showed that high FPG was a major contributor to ASDR and ASDALYR in PAD patients with a most significant upward trend.We also discovered that high FPG contributed to the highest ASDR after 85-90 y/o in males and caused a great amount of ASDR after 65-70 y/o in females.Additionally, high FPG was predicted to cause more PAD-related burden in lower SDI regions.This may be related to rising prevalence of diabetes in lower SDI regions 44 .Recent studies showed high-income countries had made substantial improvements in health care services, complication rates, and diabetes deaths 45 .Thus, diabetes management should be emphasized, especially in lower SDI regions and males aged over 85-90 y/o and females aged over 65-70 y/o.Similar to diabetes, smoking is an important PAD risk factor.Smoking cessation is associated with decreased mortality and improved amputation-free survival among symptomatic PAD patients 46,47 .Our study found that smoking was the main contributor for PAD in males, consistent with previous reports [48][49][50] .Moreover, we also found that smoking contributed to the highest ASDR before 85-90 y/o in males and accounted for the highest ASDR before 65-70 y/o in females.Our study also indicated that the attribution of smoking to disease burden is declining globally, especially in high SDI and WBI regions.A study based on GBD 2019 found that higher SDI regions implemented better tobacco restriction policies, which reduced smoking prevalence and improved population health 51 .Therefore, the public and healthcare providers should strengthen smoking management, especially in lower SDI regions and males aged before 85-90 y/o and females before 65-70 y/o.
Blood pressure control reduced cardiovascular-related risks for death and other adverse outcomes.Several population-based studies have shown that hypertension is associated with PAD 52,53 .Our studies showed that high SBP contributed to the largest ASDR and ASDALYR of PAD in females and was the third risk factor of disease burden in males.We also found that high SBP contributed to the highest ASDR in females aged over 65-70 y/o.Though high SBP led to less disease burden globally, as observed in our study, it aggravated more disease burden of PAD in low, low-middle and middle SDI and low WBI regions compared to higher SDI and WBI regions, implying that more focus is needed on SBP control in females, especially aged over 65-70 y/o, and in lower income regions 54 .
Several studies have shown an association between chronic kidney disease and PAD 55,56 , Other studies reported poorer outcomes for PAD patients with chronic kidney disease, both in limb loss and mortality [57][58][59] .According to our results, kidney dysfunction was the fourth contributor to the PAD disease burden.It was supposed to cause more PAD-related burden, especially in lower SDI regions and lower WBI level countries.Statins and angiotensin-converting enzyme (ACE) inhibitors were reported to associate with improved renal function and reduced cardiovascular risk in patients with PAD 60 .Based on current evidence, initiation of appropriate treatment (e.g., statins and ACE inhibitors) should be implemented to preserve renal function and improve PAD-related morbidity and mortality 61 .
There are still limitations to consider when interpreting the study's findings.Firstly, our analysis depended on initial GBD studies' quality, lower SDI regions are less likely to have ankle-brachial index test, and other diagnostics approaches to PAD, which may lead to data omission.Secondly, GBD does not stratify PAD based on whether affected individuals were symptomatic or asymptomatic.Nor is it clear if PAD diagnostic criteria across countries were consistent.Thirdly, as with any observational study, there are likely unmeasured confounding factors outside this analysis' scope.
In conclusion, PAD as a public health challenge has been underestimated for three decades.Higher SDI regions are seeing a decline in PAD burden, but lower SDI regions are seeing a rise.High SBP was the leading cause of PAD-related death and DALYs for females, while smoking was the leading cause for males.High FPG has become the secondary critical risk factor predicted to cause more PAD disease burden, especially in lower SDI regions.These findings helped formulate tailored strategies to alleviate disease burden.Future research should focus on developing interventions targeting risk factors to reduce PAD-related burden.

Figure 1 .
Figure 1.The global burden of peripheral arterial disease in 204 countries and territories.(a) The agestandardized incidence rate in 2019.(b) The age-standardized prevalence rate in 2019.(c) The age-standardized death rate in 2019.(d) The age-standardized DALYs rate in 2019.DALY: disability-adjusted life-year.Maps generation was performed using the open-source software R (version 4.1.0)with package of "maps" (URL of Package 'maps': https:// cran.r-project.org/ web/ packa ges/ maps/ index.html; under a GNU general public license version 2: https:// cran.r-project.org/ web/ licen ses/ GPL-2). 4 .

Figure 3 .
Figure 3.The EAPCs of PAD-related ASIR, ASPR, ASDR, ASDALYR, ASYLDR and ASYLLR in globe, SDI and WBI regions, by sex.The number indicates the corresponding EAPCs.Red indicates higher values, while blue indicates lower values of EAPCs.ASIR: age-standardized incidence rate; ASPR: age-standardized prevalence rate; ASDR: age-standardized deaths rate; ASDALYR: age-standardized disability-adjusted life-years rate; ASYLDR: age-standardized years lived with disability rate; ASYLLR: age-standardized years of life lost rate; PAD: peripheral arterial disease; EAPC: estimated annual percentage change; SDI: socio-demographic index; WBI: World Bank income level.

Figure 6 .
Figure 6.The PAFs of attributable risk factors for PAD-related ASDR in 2019 among different age groups, by sex.PAF: population attributable fraction.