jfa journal

AND option

OR option

LONG-TERM ASPIRIN USE AND SELFREPORTED WALKING SPEED IN OLDER MEN: THE PHYSICIANS’ HEALTH STUDY

 

A.R. Orkaby1,2, A.B. Dufour3, L. Yang3, H.D. Sesso2, J.M. Gaziano1,2, L. Djousse1,2, J.A. Driver1,2, T.G. Travison3

 

1. New England GRECC, VA Boston Healthcare System, Boston, MA, USA; 2. Brigham & Women’s Hospital and Harvard Medical School, Boston, MA, USA; 3. Marcus Institute for Aging Research, Hebrew Senior Life, and Harvard Medical School, Boston, MA, USA.

Corresponding Author: Ariela Orkaby, MD MPH, Assistant Professor of Medicine, Harvard Medical School, Geriatric Research, Education and Clinical Center, VA Boston Healthcare System, 150 South Huntington St, Boston, MA 02130, USA, aorkaby@bwh.harvard.edu

J Frailty Aging 2021;in press
Published online September 17, 2021, http://dx.doi.org/10.14283/jfa.2021.36

 


Abstract

Background: Mobility limitation is a component of frailty that shares a bidirectional relationship with cardiovascular disease (CVD). Data are limited on the role of established CVD prevention therapies, such as aspirin, for prevention of frailty and mobility limitation.
Objectives: Examine the association between long-term aspirin use and walking speed.
Design, Setting, Participants: Prospective cohort of 14,315 men who participated in the Physicians’ Health Study I, a completed randomized controlled trial of aspirin (1982-1988), with extended post-trial follow-up.
Measurements: Annual questionnaires collected data on aspirin use, lifestyle and other factors. Average annual aspirin use was categorized for each participant: ≤60 days/year and >60 days/year. Mobility was defined according to self-reported walking pace, categorized as: don’t walk regularly (reference), easy/casual <2mph, normal ≥2-2.9mph, or brisk/very brisk ≥3mph. Propensity scoring balanced covariates between aspirin categories. Multinomial logistic regression models estimated odds of being in each self-reported walking category.
Results: Mean age was 70±8 years; mean aspirin use 11 years. There were 2,056 (14.3%) participants who reported aspirin use ≤60 days/year. Aspirin use >60 days/year was associated with drinking alcohol, smoking, hypertension, heart disease and stroke, while ≤60 days/year was associated with anticoagulation use and bleeding history. In all, 13% reported not walking regularly, 12% walked <2 mph, 44% walked ≥2-2.9 mph, and 31% walked ≥3 mph. After propensity score adjustment, regular aspirin use was associated with a faster walking speed. Odds ratios (95% confidence intervals) were 1.16 (0.97 to 1.39), 1.24 (1.08 to 1.43), and 1.40 (1.21 to 1.63) for <2 mph, ≥2-2.9 mph and ≥3 mph, respectively, compared to not walking regularly (p-trend<0.001).
Conclusions: In this cohort of older men, long-term aspirin use is associated with a greater probability of faster walking speed later in life.

Key words: Aspirin, mobility, gait speed, frailty.


 

Introduction

One of the most feared consequences of aging is loss of function and resultant loss of independence (1). Mobility is a critical aspect of function (l) independence, and limitations in mobility – specifically as measured by usual walking speed – are associated with an increased risk of morbidity and mortality (2-4). To date, the only proven therapy to prevent mobility limitation is exercise and leisure-time activity (5).
Mobility limitation is often considered a component of frailty, a state of decreased resilience to stressors that is more common in older adults (6) We have previously demonstrated that slow walking speed is associated with cardiovascular disease (CVD) events (7), and it is thought that frailty and dismobility have a bidirectional relationship with CVD, including peripheral vascular disease (PVD) (8-10). This suggests that pharmacologic therapies targeting CVD may prevent frailty and, specifically, mobility limitation, but this has yet to be investigated.
Among such therapies, aspirin is an intriguing option as it has both anti-inflammatory and anti-thrombotic properties which can improve large and small vessel blood flow and muscle function (11-13). Data from the ASPirin in Reducing Events in the Elderly (ASPREE) trial, demonstrated a reduction in persistent ADL disability among older adults randomized to aspirin vs placebo (14). However, in older adults, data on aspirin are limited and conflicting, with increasing risks of major bleeding largely outweighing benefit for those without a history of CVD (15, 16). In prior work we have shown that long term aspirin use in men is associated with a lower risk of frailty (17). We hypothesized that long term aspirin use, especially if started in middle age when risks of aspirin are minimal, would be associated with a lower risk of mobility limitation measured according to walking speed. We used 15-year follow-up data from men enrolled in the extended observational phases of the Physician’s Health Study (PHS) I.

 

Methods

Cohort

PHS I is a completed double-blind, 2×2 factorial placebo-controlled trial that randomized 22,071 male physicians to either aspirin or placebo, and to beta-carotene or placebo, beginning in 1982 (18-20). At the start of the trial, participants were free of cancer and CVD. The aspirin intervention component of the trial ended early in 1988 at the recommendation of the PHS Data Safety Monitoring Board, due to the highly significant reduction in the rate of total myocardial infarctions among those assigned to aspirin versus placebo (19). Observational follow-up was extended after completion of the trial with annual questionnaires.
Detailed questionnaires were sent to participants annually to collect information on clinical variables, medications, and lifestyle from 1982 through 2012. Questions regarding walking speed were added to the annual questionnaire returned from 2001-2003. Participants responding to this questionnaire were eligible for the cross-sectional analyses described here, which compares participants to one another on the basis of cumulative aspirin exposure irrespective of randomized beta-carotene assignment. Of 14,896 participants who responded to the self-reported walking speed question, 581 were excluded due to missing covariate data. All participants were aged ≥58 years at the time of the walking speed assessment.
All participants executed written informed consent, and the PHS trial and subsequent follow-up were approved by the Institutional Review Board at Brigham and Women’s Hospital in Boston, MA, USA.

Exposure

For the duration of the aspirin trial, participants were randomized to 325 mg aspirin or placebo every other day. The aspirin arm of the trial was stopped after an average of 60 months of follow-up (19). Once the aspirin arm of the trial ended, treatment crossover to aspirin use among those who had been assigned to placebo was greater than 70% (21). Ongoing use of aspirin was asked on every annual questionnaire, with the following question: “Over the past 12 months, on how many days have you taken aspirin or medication containing aspirin? 0 days, 1-13 days, 14-30 days, 31-60 days, 61-90 days, 91-120 days, 121-180 days, 180+days.” For the analyses reported here, aspirin exposure was computed from responses to the questionnaires sent out 1988-2001, corresponding to the year that walking pace was added. Average annual use of aspirin was summed and categorized for each participant as follows: ≤60 days/year (low use), >61 days/year (regular use) for the follow up years post-trial, as has been done previously in PHS (22). Information on the actual dose of aspirin was not available once the trial had ended.

Outcome

A previously validated self-assessment of average walking speed was added to the questionnaire in 2001 (23, 24). The question was asked as follows: “What is your usual walking pace?” with the following possible responses: “Don’t walk regularly; Easy, casual (<2 mph); Normal, average (2-2.9mph); Brisk pace (3-3.9mph); Very brisk, striding (4mph or faster)”. After examining the data distribution, we categorized walking pace into 4 categories: don’t walk regularly, easy casual <2mph, normal ≥2-2.9mph, and brisk or very brisk ≥3mph.

Other Baseline Covariates

Information on demographics, comorbidities, and health, including history of bleeding, heart disease, stroke, peripheral artery disease, arthritis, migraine or headache, atrial fibrillation, or anticoagulation use, was drawn from all prior questionnaires to ensure complete capture of data. Smoking status was quantified as “never, past, or current”. Alcohol consumption was categorized as “daily, weekly, monthly, or rarely”. Cumulative non-aspirin non-steroidal anti-inflammatory drug use (NSAID) was quantified as the average number of years of self-reported NSAID use that were >60 days per year.

Statistical Analysis

Because of significant crossover to aspirin after completion of the intervention phase of the trial and dissolution of its randomized groupings, we developed a propensity score to balance the degree of aspirin exposure over a maximum 15 years of the post-intervention follow-up by relevant covariates. Estimation of the propensity score accounted for the initial PHS trial randomization; other variables that might be related to post-intervention aspirin use (e.g. history of bleeding, heart disease) and health-related risk factors and outcomes (e.g. smoking history). Exercise and other potential mediators or moderators of the influence of aspirin on walking speed were excluded from the propensity scoring model. We examined the distribution of propensity scores using kernel density plots to determine whether there was sufficient overlap between exposed vs unexposed (25, 26). We initially considered aspirin exposure in 2 or 3 categories of use (≤60 days vs >60 days or ≤60 days vs 61-180 vs >180 days per years). Two categories had the best overlap and were used for all analyses, as in our prior work (17, 22). Estimation employed inverse probability of treatment weighting (27-29). We compared the standardized differences between baseline covariates before and after propensity score adjustment to ensure <10% difference (25).
Descriptive characteristics were obtained after adjustment for propensity scores. Multinomial logistic regression models estimated the odds of prevalent mobility defined according to self-reported walking speed category among aspirin users relative to non-users (≤60 d/yr) while contrasting each level of walking speed to “Don’t walk regularly” category. Potential effect modification by age, history of heart disease, arthritis, bleeding and exercise frequency were examined by inspection and application of interaction tests, stratification and subgroup analysis.
Model-based estimates of association were accompanied by 95% confidence intervals (95% CI). Prespecified hypothesis testing was performed at an α=0.05 level. All analyses were conducted in SAS version 9.4 (SAS Institute, Cary, NC, USA) and R version 3.3.2 (R Foundation for Statistical Computing, Vienna, Austria).

 

Results

In total, 14,315 male, predominantly white (93%), physicians were included in this study. At time of walking speed assessment, the mean (standard deviation; SD) age was 70 (8) years with range 58-100 years; mean (SD) duration of aspirin use was 11 (5) years (range 0-18 years). Aspirin use was reported as ≤60 days/year by 2,056 (14.4%) participants. For walking speed, 13.1% reported not walking regularly, 12.2% walked <2 mph, 44.1% walked ≥2-2.9 mph, and 30.7% walked ≥3 mph.
Prior to propensity score adjustment, those who reported greater aspirin use were more likely to report drinking alcohol daily, prior smoking, and a history of atrial fibrillation, diabetes, hypertension, coronary heart disease (CHD), and stroke. Lower frequency of aspirin use was associated with greater anticoagulation use and bleeding history such as gastritis and melena. Following propensity score IPTW adjustment, there were no significant differences between groups (Table 1). With adjustment for propensity score, we found a graded association between regular aspirin use and faster walking speed. The multiplicative increase (95% CI) attributable to high aspirin use in odds of faster walking speed as compared to not regularly walking was 1.16 (0.97 to 1.39), 1.24 (1.08 to 1.43), and 1.40 (1.21 to 1.63) for those reporting walking speeds of <2 mph, ≥2-2.9 mph and ≥3 mph, respectively, (p-trend<0.001) (Table 2). In a sensitivity analysis, after adjusting for the propensity score in the crude model, results were unchanged.

Table 1. Cohort characteristics of 14,315 PHS participants according to average annual aspirin use, before and after propensity score adjustment

Table 2. Association between regular aspirin use and self-reported walking pace in 14,315 PHS participants before and after propensity score weighting

Legend: Sample sizes are provided for each category to allow for calculation of the raw ORs. For example, comparing “normal” walkers with those who “don’t walk” the estimated unadjusted OR is 1.18, as shown in the table. This is interpreted as: for an individual in the high-aspirin group, the relative odds of being in the “normal» walking group than the «don’t walk” group are18% greater with high aspirin exposure vs low aspirin exposure.

 

Subgroup analyses generally showed little evidence of differential effects across pre-specified categories of risk factors, including age, arthritis, bleeding or exercise frequency. The only subgroup in which there was a statistically significant interaction was among those with history of CHD (p<0.0008). The OR (95% CI) for those with prior CHD and high aspirin use (vs low) was 1.15 (0.80 to 1.64), 1.72 (1.26 to 2.35), and 2.84 (1.93 to 4.16) for those reporting <2 mph, ≥2-2.9 mph and 3+ mph, respectively, compared to 1.18 (0.96 to 1.45), 1.15 (0.98 to 1.34), and 1.24 (1.05 to 1.46) for those without CHD. (Figures 1 and 2).

Figure 1. Forest plot showing the relative odds of increased mobility according to level of aspirin use, stratifying by age and and history of heart disease. The referent mobility group is those who do not engage in regular walking

Figure 2. Forest plot showing the relative odds of increased mobility according to level of aspirin, stratifying on history of arthritis, bleeding, and weekly exercise. The referent mobility group is those who do not engage in regular walking

 

Discussion

In this 15-year post-trial follow up study of 14,315 male physicians, long term aspirin use, started in middle age, was associated with a greater probability of faster walking speed in late life. Our results suggest that decreased walking speed might be prevented through regular aspirin use, independent of medical history and health behaviors. Importantly, however, results suggest that after consideration of these factors, the associations due to aspirin exposure are more strongly manifested among older individuals with a history of heart disease. This observation is in line with the hypothesis that those cardiovascular morbidity and frailty have shared underlying pathophysiology and may be targeted by regular aspirin use.
Aspirin lowers the risk of cardiovascular events by both preventing platelet aggregation and reducing inflammation. In patients with PVD, aspirin is an established part of the treatment plan, either alone or in combination with other antiplatelets or anticoagulants (30, 31). However, clinical trial data has not consistently demonstrated a benefit of aspirin of intermittent claudication. Although aspirin has anti-inflammatory and vasodilatory properties which might improve walking speed, prior clinical data in those with PVD does not support this hypothesis. While most older adults do not have clinical PVD, the microvascular changes that occur with aging contribute to decreased blood flow to leg muscles and may explain in part the natural slowing in walking and increased sarcopenia that is seen with physiologic aging (3, 32). We hypothesize that regular use of aspirin over years, in late middle age and early old age, may improve blood flow and circulation to distal muscles particularly in those with CVD, and therefore lead to less mobility limitation in later decades, in those with heart disease in particular. This is likely mediated by the vasodilatory and anti-inflammatory properties of aspirin.Randomized trials of aspirin for mobility are limited. The ASPREE trial examined the effect of aspirin on disability-free survival, CV events, and mortality in adults aged 70 and older, over 5-years (15, 33, 34). In secondary analyses there was a signal for a lower rate of persistent physical disability (HR 0.85, 95% CI 0.70 to 1.03) (33). and possible benefit for mobility. However, there are serious concerns regarding the safety of aspirin that need to be considered, particularly related to fatal bleeding events (15). It is possible that any benefit of aspirin for prevention of mobility limitation would need to begin earlier in life, however we found that the association between long-term use of aspirin and lower risk of slow walking remained even after accounting for age.
In subgroup analyses we explored whether aspirin use is more effective in certain disease states which may be impacted by aspirin use. Although overall results were similar, there was a statistically significant interaction for those with a history of CHD and a non-significant suggestion of greater benefit among those who do not exercise regularly. Cardiovascular disease has multiple causes, including increased inflammation and thrombotic milieu. That those with CHD had evidence of greater benefit from aspirin may suggest that those using aspirin for secondary prevention of CVD might improve walking speed later in life by reducing their burden of CVD over the lifespan by preventing future CV events and resultant disability. On the other hand, those who exercise may have sufficient benefit from improving blood flow to extremities (35) that the addition of aspirin does not further improve walking. Exercise has been shown to be one of the few modalities that can improve function and frailty at all stages (36), and it is possible that the benefits of aspirin are most beneficial in those who are inactive or who have biologically active CVD.
There are several important limitations to consider. First, this cohort is entirely male and largely made up of individuals identifying as white within a US context, and the study should be repeated in a cohort of women. However, women are also more likely to outlive men by 6-8 years and data that is sex specific is needed in aging research. Second, walking speed was assessed using self-report which could either over or under estimate actual walking pace, although the question used has been well validated (23, 24). Although the end-point mixes aspects of physical activity, function, and ability or disability, we interpret this as an important measure of actualized functional capacity in the real-world setting (21). Third, even though we used propensity score methods to address confounding by indication, reverse causality remains possible. Fourth, information on dose of aspirin was not available after completion of the trial and we are unable to examine the role of aspirin dose on the outcome. Future work to understand the relationship between aspirin and mobility could include plasma analysis for inflammatory biomarkers and mediation analysis to understand what role aspirin may play in reducing inflammation and preventing frailty.
In conclusion, in this cohort of older men, we found that long-term aspirin use started in late midlife was associated with lower odds of mobility limitation, as defined by self-reported walking speed. Future work should replicate these findings in women and further explore the protentional role of aspirin as a preventive strategy for frailty and mobility limitation.

 

Acknowledgments: The authors gratefully acknowledge the patients and research staff who participated in the Physicians’ Health Study. We are grateful for Natalya Gomelskaya for her statistical support. All authors contributed to study conception, writing, and editing the manuscript. L.Y., A.B.D., and T.G.T. conducted the statistical analysis. The sponsor had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.

Funding: This work was supported by a Career Development award to A.R.O. from the Boston Claude D. Pepper Older Americans Independence Center, National Institute on Aging grant P30-AG013679. A.R.O. is also supported by Veterans Affairs CSR&D CDA-2 IK2-CX001800 and National Institute on Aging grant R03-AG060169. The Physicians’ Health Study is funded by grants CA-34944, CA-40360, and CA-097193 from the National Cancer Institute and grants HL-26490 and HL-34595 from the National Heart, Lung, and Blood Institute, Bethesda, MD.

Conflicts of Interest: J.M.G. reports serving as a consultant and receiving honoraria for speaking for Bayer.

Ethical standards: All participants provided written informed consent. The trial was approved by the Institutional Review Board at Brigham and Women’s Hospital.

 

References

1. Parry SW, Finch T and Deary V. How should we manage fear of falling in older adults living in the community? BMJ (Clinical research ed). 2013;346:f2933. doi: 10.1136/bmj.f2933
2. Studenski S, Perera S, Patel K, Rosano C, Faulkner K, Inzitari M, Brach J, Chandler J, Cawthon P, Connor EB, Nevitt M, Visser M, Kritchevsky S, Badinelli S, Harris T, Newman AB, Cauley J, Ferrucci L and Guralnik J. Gait speed and survival in older adults. Jama. 2011;305:50-8. doi: 10.1001/jama.2010.1923
3. Fritz S and Lusardi M. White paper: «walking speed: the sixth vital sign». Journal of geriatric physical therapy (2001). 2009;32:46-9.
4. Middleton A, Fritz SL and Lusardi M. Walking speed: the functional vital sign. Journal of aging and physical activity. 2015;23:314-22. doi: 10.1123/japa.2013-0236
5. Brown CJ and Flood KL. Mobility limitation in the older patient: a clinical review. Jama. 2013;310:1168-77. doi: 10.1001/jama.2013.276566
6. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, Seeman T, Tracy R, Kop WJ, Burke G and McBurnie MA. Frailty in older adults: evidence for a phenotype. The journals of gerontology Series A, Biological sciences and medical sciences. 2001;56:M146-56. doi: 10.1093/gerona/56.3.M146
7. Imran TF, Orkaby A, Chen J, Selvaraj S, Driver JA, Gaziano JM and Djoussé L. Walking pace is inversely associated with risk of death and cardiovascular disease: The Physicians’ Health Study. Atherosclerosis. 2019;289:51-56. doi: 10.1016/j.atherosclerosis.2019.08.001
8. Afilalo J, Alexander KP, Mack MJ, Maurer MS, Green P, Allen LA, Popma JJ, Ferrucci L and Forman DE. Frailty assessment in the cardiovascular care of older adults. Journal of the American College of Cardiology. 2014;63:747-62. doi: 10.1016/j.jacc.2013.09.070
9. Bouillon K, Batty GD, Hamer M, Sabia S, Shipley MJ, Britton A, Singh-Manoux A and Kivimaki M. Cardiovascular disease risk scores in identifying future frailty: the Whitehall II prospective cohort study. Heart (British Cardiac Society). 2013;99:737-42. doi: 10.1136/heartjnl-2012-302922
10. Newman AB, Gottdiener JS, McBurnie MA, Hirsch CH, Kop WJ, Tracy R, Walston JD and Fried LP. Associations of subclinical cardiovascular disease with frailty. The journals of gerontology Series A, Biological sciences and medical sciences. 2001;56:M158-66. doi: 10.1093/gerona/56.3.M158
11. Vane JR and Botting RM. The mechanism of action of aspirin. Thrombosis research. 2003;110:255-8. doi: 10.1016/S0049-3848(03)00379-7
12. Oh J, Sinha I, Tan KY, Rosner B, Dreyfuss JM, Gjata O, Tran P, Shoelson SE and Wagers AJ. Age-associated NF-kappaB signaling in myofibers alters the satellite cell niche and re-strains muscle stem cell function. Aging. 2016;8:2871-2896. doi: 10.18632/aging.101098
13. Morris T, Stables M, Hobbs A, de Souza P, Colville-Nash P, Warner T, Newson J, Bellingan G and Gilroy DW. Effects of low-dose aspirin on acute inflammatory responses in humans. Journal of immunology (Baltimore, Md : 1950). 2009;183:2089-96. doi: 10.4049/jimmunol.0900477
14. Woods RL, Espinoza S, Thao LTP, Ernst ME, Ryan J, Wolfe R, Shah RC, Ward SA, Storey E, Nelson MR, Reid CM, Lockery JE, Orchard SG, Trevaks RE, Fitzgerald SM, Stocks NP, Williamson JD, McNeil JJ, Murray AM and Newman AB. Effect of Aspirin on Activities of Daily Living Disability in Community-Dwelling Older Adults. The journals of gerontology Series A, Biological sciences and medical sciences. 2020. doi: 10.1093/gerona/glaa316
15. McNeil JJ, Wolfe R, Woods RL, Tonkin AM, Donnan GA, Nelson MR, Reid CM, Lockery JE, Kirpach B, Storey E, Shah RC, Williamson JD, Margolis KL, Ernst ME, Abhayaratna WP, Stocks N, Fitzgerald SM, Orchard SG, Trevaks RE, Beilin LJ, Johnston CI, Ryan J, Radziszewska B, Jelinek M, Malik M, Eaton CB, Brauer D, Cloud G, Wood EM, Mahady SE, Satterfield S, Grimm R and Murray AM. Effect of Aspirin on Cardiovascular Events and Bleeding in the Healthy Elderly. The New England journal of medicine. 2018;379:1509-1518. doi: 10.1056/NEJMoa1805819
16. Bibbins-Domingo K. Aspirin Use for the Primary Prevention of Cardiovascular Disease and Colorectal Cancer: U.S. Preventive Services Task Force Recommendation Statement. Annals of internal medicine. 2016;164:836-45. doi: 10.7326/M16-0577
17. Orkaby AR, Yang L, Dufour AB, Travison TG, Sesso HD, Driver JA, Djousse L and Gaziano JM. Association Between Long-Term Aspirin Use and Frailty in Men: The Physicians’ Health Study. The journals of gerontology Series A, Biological sciences and medical sciences. 2020. doi: 10.1093/gerona/glaa233
18. Findings from the aspirin component of the ongoing Physicians’ Health Study. The New England journal of medicine. 1988;318:262-4. doi: 10.1056/NEJM198801283180431
19. Final report on the aspirin component of the ongoing Physicians’ Health Study. Steering Committee of the Physicians’ Health Study Research Group. The New England journal of medicine. 1989;321:129-35. doi: 10.1056/NEJM198907203210301
20. Hennekens CH, Buring JE, Manson JE, Stampfer M, Rosner B, Cook NR, Belanger C, LaMotte F, Gaziano JM, Ridker PM, Willett W and Peto R. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease. The New England journal of medicine. 1996;334:1145-9. doi: 10.1056/NEJM199605023341801
21. Sturmer T, Glynn RJ, Lee IM, Manson JE, Buring JE and Hennekens CH. Aspirin use and colorectal cancer: post-trial follow-up data from the Physicians’ Health Study. Annals of internal medicine. 1998;128:713-20. doi:10.7326/0003-4819-128-9-199805010-00003
22. Driver JA, Logroscino G, Lu L, Gaziano JM and Kurth T. Use of non-steroidal anti-inflammatory drugs and risk of Parkinson’s disease: nested case-control study. BMJ (Clinical research ed). 2011;342:d198. doi: 10.1136/bmj.d198
23. Wolf AM, Hunter DJ, Colditz GA, Manson JE, Stampfer MJ, Corsano KA, Rosner B, Kriska A and Willett WC. Reproducibility and validity of a self-administered physical activity questionnaire. International journal of epidemiology. 1994;23:991-9. doi: 10.1093/ije/23.5.991
24. Hu FB, Sigal RJ, Rich-Edwards JW, Colditz GA, Solomon CG, Willett WC, Speizer FE and Manson JE. Walking compared with vigorous physical activity and risk of type 2 diabetes in women: a prospective study. Jama. 1999;282:1433-9. doi: 10.1001/jama.282.15.1433
25. Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Statistics in medicine. 2009;28:3083-107. doi: 10.1002/sim.3697
26. Crump RK, Hotz VJ, Imbens GW and Mitnik OA. Dealing with limited overlap in estimation of average treatment effects. 2009;96:187-199. doi: 10.1093/biomet/asn055
27. Elze MC, Gregson J, Baber U, Williamson E, Sartori S, Mehran R, Nichols M, Stone GW and Pocock SJ. Comparison of Propensity Score Methods and Covariate Adjustment: Evaluation in 4 Cardiovascular Studies. Journal of the American College of Cardiology. 2017;69:345-357. doi: 10.1016/j.jacc.2016.10.060
28. Curtis LH, Hammill BG, Eisenstein EL, Kramer JM and Anstrom KJ. Using inverse probability-weighted estimators in comparative effectiveness analyses with observational databases. Medical care. 2007;45:S103-7. doi: 10.1097/MLR.0b013e31806518ac
29. Kim DH, Pieper CF, Ahmed A and Colon-Emeric CS. Use and Interpretation of Propensity Scores in Aging Research: A Guide for Clinical Researchers. Journal of the American Geriatrics Society. 2016;64:2065-2073. doi: 10.1111/jgs.14253
30. Gerhard-Herman MD, Gornik HL, Barrett C, Barshes NR, Corriere MA, Drachman DE, Fleisher LA, Fowkes FG, Hamburg NM, Kinlay S, Lookstein R, Misra S, Mureebe L, Olin JW, Patel RA, Regensteiner JG, Schanzer A, Shishehbor MH, Stewart KJ, Treat-Jacobson D and Walsh ME. 2016 AHA/ACC Guideline on the Management of Patients With Lower Extremity Peripheral Artery Disease: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2017;135:e686-e725. doi: 10.1161/CIR.0000000000000470
31. Hess CN and Hiatt WR. Antithrombotic Therapy for Peripheral Artery Disease in 2018. Jama. 2018;319:2329-2330. doi: 10.1001/jama.2018.5422
32. Proctor DN, Shen PH, Dietz NM, Eickhoff TJ, Lawler LA, Ebersold EJ, Loeffler DL and Joyner MJ. Reduced leg blood flow during dynamic exercise in older endurance-trained men. J Appl Physiol (1985). 1998;85:68-75. doi: 10.1152/jappl.1998.85.1.68
33. McNeil JJ, Woods RL, Nelson MR, Reid CM, Kirpach B, Wolfe R, Storey E, Shah RC, Lockery JE, Tonkin AM, Newman AB, Williamson JD, Margolis KL, Ernst ME, Abhayaratna WP, Stocks N, Fitzgerald SM, Orchard SG, Trevaks RE, Beilin LJ, Donnan GA, Gibbs P, Johnston CI, Ryan J, Radziszewska B, Grimm R and Murray AM. Effect of Aspirin on Disability-free Survival in the Healthy Elderly. The New England journal of medicine. 2018;379:1499-1508. doi: 10.1056/NEJMoa1800722
34. McNeil JJ, Nelson MR, Woods RL, Lockery JE, Wolfe R, Reid CM, Kirpach B, Shah RC, Ives DG, Storey E, Ryan J, Tonkin AM, Newman AB, Williamson JD, Margolis KL, Ernst ME, Abhayaratna WP, Stocks N, Fitzgerald SM, Orchard SG, Trevaks RE, Beilin LJ, Donnan GA, Gibbs P, Johnston CI, Radziszewska B, Grimm R and Murray AM. Effect of Aspirin on All-Cause Mortality in the Healthy Elderly. The New England journal of medicine. 2018;379:1519-1528. doi: 10.1056/NEJMoa1803955
35. Hildebrandt W, Schwarzbach H, Pardun A, Hannemann L, Bogs B, König AM, Mahnken AH, Hildebrandt O, Koehler U and Kinscherf R. Age-related differences in skeletal muscle microvascular response to exercise as detected by contrast-enhanced ultrasound (CEUS). PLoS One. 2017;12:e0172771. doi: 10.1371/journal.pone.0172771
36. Clegg A, Young J, Iliffe S, Rikkert MO and Rockwood K. Frailty in elderly people. Lancet. 2013;381:752-62. doi: 10.1016/S0140-6736(12)62167-9