Physical Activity in Patients with Symptomatic Peripheral Artery Disease: Insights from the PORTRAIT Registry

Abstract

Objective:

A physically active lifestyle reduces the risk of cardiovascular events and functional impairment in patients with peripheral artery disease (PAD). There are limited data on the patterns of physical activity in patients with PAD compared across countries.

Methods:

Self-reported physical activity (sedentary vs not) was obtained at enrolment, 3, 6, and 12 months in the US and Netherlands’ cohorts of the Patient-centered Outcomes Related to Treatment Practices in Peripheral Arterial Disease: Investigating Trajectories (PORTRAIT) registry of patients with new or worsening claudication. Multivariable repeated measures using modified Poisson regression analysis compared the proportion of sedentary participants over time between countries and to identify factors that attenuate intercountry differences.

Results:

Of 1 098 participants, 743 (67.7%) and 355 (32.3%) were recruited from the USA and the Netherlands respectively. Compared to the Netherlands, participants from the US were older (mean age 68.6 vs 65.3 years; p < .001), more obese (41.3% vs 20.5%; p < .001), and more likely to be female (41.3% vs 31.4%; p = .002). There were fewer current smokers (30.1% vs 52.8%; p < .001) and supervised exercise referrals (1.6% vs 63.9%; p < .001) in the US compared with the Netherlands. US participants were more sedentary at baseline (43.7% vs. 34.1%; p < .001). Sedentary behaviour decreased after 3 months in both countries, then diverged with an increase in sedentary participants in the USA. Risk of sedentary behaviour was significantly greater in the USA compared with the Netherlands at 12 months, after adjustment of sociodemographic, lifestyle factors, and comorbidities (relative risk [RR] 1.56, 95% confidence interval [CI] 1.08 – 2.25; p = .020) but was attenuated after accounting for referral to supervised exercise (RR 1.20, 95% CI 0.67 – 2.16; p =.54).

Conclusion:

Referral to supervised exercise was key in explaining the observed difference in the physical activity levels between patients with PAD in the USA and the Netherlands. Further promotion of supervised exercise for PAD may improve physical activity in patients with PAD and modify cultural norms of inactivity in the US.

INTRODUCTION

Sedentary lifestyles, a known risk factor for cardiometabolic disease and adverse outcomes,^1,2 appear to be on the rise worldwide,³ and specifically among women with cardiovascular disease.⁴ Given the increased risk of death, myocardial infarction, stroke, and functional decline and worse outcomes associated with low levels of physical activity,^5,6 it is important to understand activity patterns, particularly in populations at higher risk and for whom physical activity can improve outcomes. Patients with symptomatic peripheral artery disease (PAD) constitute a high-risk cohort of patients who are at a particularly increased risk for cardiovascular events and mortality.⁷ These patients also experience significant physical limitations due to claudication symptoms that can negatively influence their overall physical activity levels, and increase the risk of sedentary behaviour. Consequently, treatment of PAD focuses not only on cardiovascular event reduction, but also on improving symptoms and functional limitations related to PAD.⁸

There are a number of ways in which patients’ routine activity might vary. For example, the use of supervised exercise programmes can increase exercise tolerance and reduce cardiovascular risk in patients with PAD.⁹ It is also possible that cultural “norms” and environmental factors in different countries can influence patients’ activity.¹⁰ In some countries, walking and public transportation are much more common than in others. Understanding these differences can help identify patient and treatment characteristics that might be exploited to improve activity in all patients with PAD.

To date, no study has tracked physical activity over time in patients with symptomatic PAD in a real-world setting and how this varies across different countries. To address this gap in knowledge, data were leveraged from an international, prospective registry to assess for changes in leisure time physical activity level in patients with PAD and newly diagnosed intermittent claudication or deterioration of previously diagnosed intermittent claudication. Patient and treatment characteristics that might modify these differences were investigated to provide insights into novel strategies for improving the care and outcomes of patients with PAD. An implication of this study would be to detect disparities in aspects of care for patients with PAD between countries, which may help guide the allocation of resources to key disease-management strategies or risk factors.

METHODS

Study design and participants

The Patient-centered Outcomes Related to Treatment Practices in Peripheral Arterial Disease: Investigating Trajectories (PORTRAIT) study is a prospective, international observational registry of patients evaluated for new or worsening claudication at 16 vascular specialty care clinics across the USA, the Netherlands, and Australia. For this analysis, participants in the Australian cohort were excluded as they had only one participating site, limiting the generalisability of the findings to that country. The design and methodology of the PORTRAIT study has been previously described in detail.¹¹ Briefly, consecutive patients with symptomatic PAD (resting ankle brachial index [ABI] of ≤ 0.90 or a significant drop in postexercise ankle pressure of ≥ 20 mmHg) were enrolled in vascular specialty clinics between June 2011 and December 2015. A baseline interview was performed by trained research staff to obtain information about the patient’s medical history and cardiovascular lifestyle factors, as well as socioeconomic, psychosocial, and health status. Demographic information, medical history, comorbidities, and PAD diagnostic information were abstracted from the patients’ medical records. After initial evaluation, the primary treatment strategy was categorised as initial invasive treatment or non-invasive treatment (supervised exercise and medical therapy), based on receipt of a PAD intervention in the first three months after their initial PAD workup (Supplementary Figure 1) Referral to supervised exercise was independently adjudicated by trained research personnel via a detailed review of each patient’s chart. Trained data collectors conducted three, six, and 12 month telephone interviews to obtain information about physical activity.

Leisure time physical activity

To assess patients’ physical activity, the self-reported items from the INTERHEART study were used.^12,13 Leisure time physical activity was graded from 1 to 3. Grade 1 was defined as mainly sedentary, grade 2 as mild activity (e.g., yoga or easy walking), and grade 3 as moderate or strenuous exercise, including running, bicycling, and swimming.¹⁴ For this analysis, activity was dichotomised as sedentary (grade 1) or non-sedentary (≥ grade 2). Previous studies have shown that physical activity assessed by this questionnaire independently predicts all cause and cardiovascular death, and this questionnaire has been incorporated into the INTERHEART risk score.^12,13

Ethical approval statement

All study participants provided informed consent and institutional review board approval was obtained at each participating centre.

Statistical analysis

Baseline characteristics were compared between patients enrolled in the USA, and those from the Netherlands using the Student’s t test or chi square test, as appropriate. To compare the sedentary activity over time across and within countries, hierarchical (site as a random effect), repeated measures mixed effect modified Poisson models were used. These models included a time by country interaction with the primary focus on the 12 month follow up. A modified Poisson regression model was used in order to obtain relative risks, as the prevalence of physical activity was not rare, and thus odds ratios from logistic regression models would not estimate relative risks well.

To further explore variables that could explain any noted differences, demographic factors (age, sex, and race), socio-economic (education), lifestyle factors (smoking and alcohol use), comorbidities (lung disease, obesity, heart failure, ABI, and hypertension) and referral to supervised exercise were adjusted for. Invasive treatment status is not adjudicated in the Netherlands cohort; therefore, the effect of invasive treatment on physical activity in the Netherlands could not be accounted for. To assess whether invasive treatment was associated with physical activity in the larger US cohort, an unadjusted Poisson regression was conducted for the association of invasive treatment with the risk of being sedentary at 12 months. Statistical analysis was completed using SAS version 9.4 software (StatSoft, Cary, NC, USA). Two-sided p values < .05 were considered to be statistically significant.

RESULTS

Among 1 275 patients in PORTRAIT, 1 180 were enrolled in the USA or the Netherlands. Those with complete baseline and at least one follow up assessment of leisure time physical activity (n = 1 098) were included. Baseline characteristics were largely similar between the included and excluded patients (Supplementary Appendix Table 1). In the analytic cohort, patients from the USA were older, more likely to be female, had more high school graduates, had more obesity related comorbidities, worse PAD specific health status, and were less likely to be current smokers than patients from the Netherlands (Table 1). The proportion of patients referred to supervised exercise therapy was lower in the USA (1.6%) than in the Netherlands (63.9%; p < .001). There was a statistically significantly higher proportion of sedentary participants in the USA (43.7%) compared with the Netherlands (36.1%) at baseline (p < .001). Univariate predictors of sedentary behaviour at baseline are presented in Supplementary Table 2, and include female sex, race (white), country (USA), obesity, lung disease, heart failure, coronary artery disease, sleep apnoea, severe claudication, not being referred to supervised exercise, and beta blocker use.

Table 1.

Baseline characteristics of 1 098 patients with claudication evaluated at vascular specialty clinics in the United States (US) and Netherlands as part of the Patient-centered Outcomes Related to Treatment Practices in Peripheral Arterial Disease: Investigating Trajectories (PORTRAIT) registry between June 2011 and December 2015.

Characteristics	Country		p value
	USA (n = 743)	The Netherlands (n = 355)
Age — y	68.6 ± 9.6	65.3 ± 8.6	< .001
Male sex	436 (58.7)	243 (68.5)	.002
Race (white)	535 (72.0)	355 (100)	< .001
High school education or greater	634 (85.3)	152 (42.8)	< .001
Working (full/part time)	168 (22.6)	93 (26.2)	.21
Disease severity
ABI	0.67 ± 0.19	0.66 ± 0.18	< .001
Function (Rutherford category)			< .001
Mild claudication	148 (20.0)	94 (26.5)
Moderate claudication	375 (50.7)	181 (51.0)
Severe claudication	216 (29.2)	70 (19.7)
Smoking status			< .001
Never smoker	80 (10.8)	11 (3.1)
Former smoker	439 (59.1)	156 (44.1)
Current smoker	224 (30.1)	187 (52.8)
Comorbidities
Chronic lung disease	118 (15.9)	73 (20.6)	.055
Diabetes	285 (38.4)	75 (21.1)	<.001
Hypertension	666 (89.6)	218 (61.4)	<.001
Obese (BMI > 30)	297 (40.0)	35 (9.9)	<.001
CAD	401 (54.0)	98 (27.6)	<.001
CKD	112 (15.1)	16 (4.5)	<.001
CHF	111 (14.9)	9 (2.5)	<.001
Sleep apnoea	83 (11.2)	7 (2.0)	<.001
Medications
Statin	579 (78.0)	195 (55.4)	<.001
Aspirin	606 (81.7)	98 (27.8)	<.001
Beta blocker	476 (64.2)	123 (34.9)	<.001
ACEi/ARB	515 (69.4)	146 (41.5)	<.001
Supervised exercise referral	12 (1.6)	227 (63.9)	<.001

Data are presented as mean ± standard deviation or n (%). ABI = ankle brachial index; BMI = body mass index; CAD = coronary artery disease; CHF = congestive heart failure; ACEi/ARB = angiotensin converting enzyme inhibitor/angiotensin receptor blocker.

Physical activity over time

There was a decline in the proportion of patients who were sedentary from baseline to 12 months of follow up in the overall (from 40.6% to 31.1%, p_trend < .001; Fig. 1), US (from 43.7% to 36.0%, p_trend =. 11) and Netherlands (from 34.1% to 20.9%, p_trend < .001) cohorts. In the US, the proportion of sedentary patients decreased from baseline, reaching the lowest level (26.8%) at three months followed by increases to 31.2% at six months and 36% by 12 months. In the Netherlands, the proportion of sedentary participants was similar from baseline to three months (from 34.1% to 29.9%; p = .17), but decreased from three to six months (from 29.9% to 24.1%; p < .001) and from six to 12 months (from 29.9% to 20.9%; p < .001) (Fig. 1).

Figure 1.

(A) Proportion of sedentary patients in follow up out of 1 908 patients with claudication and (B) patients enrolled in vascular specialty clinics from the United States (US; n = 743) and Netherlands (n = 355) in the Patient-centered Outcomes Related to Treatment Practices in Peripheral Arterial Disease: Investigating Trajectories (PORTRAIT) registry. The p values represent the chi square for between country difference.

Adjusted risk of sedentary behaviour

The risk of sedentary behaviour was significantly greater in the USA compared with the Netherlands at the 12 month follow up after accounting for sociodemographic factors (relative risk [RR] 1.96, 95% confidence interval [CI] 1.35 – 2.83; p < .001), lifestyle factors (RR 1.77, 95% CI 1.30 – 2.41; p < .001) and comorbidities (RR 1.56, 95%CI 1.08 – 2.25; p = .02). However, when referral to supervised exercises was included in the model, the difference in sedentary behaviour between the countries at 12 months was completely attenuated (RR 1.20, 95% CI 0.67 – 2.16; p = .54 [Table 2]). There was no significant association between invasive treatment and risk of being sedentary at 12 months in the US cohort (RR 0.99, 95% CI 0.72 – 1.36; p = .94).

Table 2.

Relative risk (RR) ratios comparing the risk of sedentary behaviour among 1 908 patients with claudication at 3, 6, and 12 months after enrolment in the Patient-centered Outcomes Related to Treatment Practices in Peripheral Arterial Disease: Investigating Trajectories (PORTRAIT) registry from the United States (US; n = 743) and the Netherlands (n = 355) between June 2011 and December 2015.

Models	RR (95% CI)	p value
M1
3 mo: US vs Netherlands	1.03 (0.72–1.46)	.88
6 mo: US vs Netherlands	1.43 (1.00–2.05)	.051
12 mo: US vs Netherlands	1.87 (1.34–2.60)	< .001
M2 = M1 + sociodemographic
3 mo: US vs Netherlands	1.03 (0.71–1.50)	.88
6 mo: US vs Netherlands	1.44 (0.97–2.15)	.070
12 mo: US vs Netherlands	1.96 (1.35–2.83)	.006
M3 = M2 + lifestyle factors
3 mo: US vs Netherlands	1.09 (0.77–1.55)	.63
6 mo: US vs Netherlands	1.31 (0.90–1.89)	.15
12 mo: US vs Netherlands	1.77 (1.30–2.41)	.002
M4 = M3 + comorbidities
3 mo: US vs. Netherlands	1.07 (0.77–1.49)	.67
6 mo: US vs Netherlands	1.51 (1.03–2.21)	.036
12 mo: US vs Netherlands	1.56 (1.08–2.25)	.019
M5 = M4 + supervised exercise
3 mo: US vs Netherlands	0.57 (0.16–2.04)	.39
6 mo: US vs Netherlands	1.01 (0.54–1.90)	.97
12 mo: US vs Netherlands	1.20 (0.67–2.16)	.54

CI = confidence interval; M1 = month, country, month × country; M2 = M1 + age, sex, race, education; M3 = M2 + smoking, alcohol use; M4 = M3 + lung disease, obesity, heart failure, ankle brachial index, hypertension; M5 = M4 + referral to supervised exercise.

DISCUSSION

Many factors contribute to routine physical activity, including personal motivation, environmental factors, and cultural norms.¹⁰ However, participation in regular physical activity is an important part of the prevention and management of atherosclerotic disease, and physical activity during leisure time is a strong predictor of cardiovascular outcomes.¹⁵ As a result, major guidelines strongly recommend regular physical activity for patients with PAD, as part of a comprehensive management strategy.¹⁶ While patients with PAD can have significant limitations in their daily activity due to claudication, little is known about the evolution of physical activity over time among patients with PAD in a real world cohort. In this analysis examining physical activity over time, it was found that the proportion of patients who were mainly sedentary was significantly greater in the USA, and while there was a transient improvement in the US participants, by 12 months it was again much greater than that observed in the Netherlands. These differences were not explained by sociodemographic factors or comorbidities, but were completely attenuated after accounting for referral to supervised exercise. These findings suggest that, while there may be a cultural proclivity to be more active in the Netherlands than in the USA, a greater reliance on supervised exercise therapy may account for this and prove to be a valuable strategy in improving the lifestyles of US patients with PAD.

These unique insights into the activity of patients after a new diagnosis of claudication or worsening of chronic claudication supplements the existing literature by highlighting the importance of activity in PAD. Among persons with no prior history of PAD, lower levels of leisure time physical activity are associated with a higher risk of the future development of PAD, suggesting that physical activity may be important in the prevention of PAD.¹⁷ Among patients who have already developed PAD, low physical activity is associated with increased risk of ischaemic heart disease events.^18,19 Increased physical activity in the form of supervised exercise is a potent treatment for lower extremity PAD, with demonstrated reductions in mortality and improvement in quality of life.²⁰ In addition, claudication treatments are specifically aimed at relieving symptoms that would allow patients to increase their physical functioning. Despite the fact that all patients were being seen by a vascular specialist, fewer patients in the US were active a year after their work-up relative to where the Dutch patient cohort was with their physical activity. The importance of physical activity in this patient population has been recognised, and has prompted calls to use physical activity as end points in the evaluation of treatment modalities in PAD.²¹

Prior studies have shown that, in the general population, physical activity levels are higher among individuals in the Netherlands versus those from the USA.²² However, it remained unclear if such patterns would hold among patients with PAD, who tend to have significant physical limitations. Factors accounting for the differential physical activity between the US and the Netherlands are likely multifactorial, including differences in age and obesity.²³ In this study, the patients recruited from the Netherlands were, overall, younger and generally had lower rates of important comorbidities that may limit physical activity participation such as heart failure, lung disease, and obesity, although even after adjusting for these differences much more sedentary lifestyles were observed in the USA. Importantly, it was found that the differences between the USA and the Netherlands were completely eradicated after accounting for supervised exercise therapy referral, which was a covered health benefit in the Netherlands but not in the US during the enrolment period of PORTRAIT. The Center of Medicare and Medicaid services commenced coverage for supervised exercise in the US in 2017,²⁴ and the impact of this change will need to be assessed in future studies. Given that guidelines recommend regular physical activity for patients with PAD,^8,16 the present study suggests that there remains more room for improvement in increasing physical activity levels, especially among the US cohort.

These findings should be interpreted in the context of several potential limitations. Firstly, participation of supervised exercise therapy during follow up could not be accounted for because adjudication across all sites is still ongoing. Furthermore, invasive treatment could not be included in the models because adjudication is still ongoing for the Netherlands cohort. However, an unadjusted logistic regression found no significant relationship between receipt of invasive treatment and 12 month physical activity status in the US cohort, suggesting invasive treatment may not play a significant role in determining 12 month activity status. Secondly, physical activity was assessed using patient reported questionnaires, which may be subject to recall bias.²⁵ Additionally, how many minutes per day or week that participants were sedentary could not be determined. Despite these limitations, the questionnaire used in this study has demonstrated prognostic properties in patients with atherosclerosis and are incorporated into important and widely used cardiovascular risk prediction scores.²⁶ Thirdly, the observational nature of this study precludes the elimination of biases due to residual measured or unmeasured confounders. In particular, the marked differences in the use of supervised exercise therapy referral between the USA and the Netherlands may serve as a “marker” of reduced sedentary activity, rather than a mediator of these differences.

Conclusion

Patients seeking care for symptomatic PAD at specialty clinics in the USA and the Netherlands were less sedentary over time. Patients in the USA were observed to be more sedentary at baseline and during follow up than patients in the Netherlands. This difference in physical activity over time between the countries was not explained by differences in the distribution of demographic characteristics and comorbidities but was markedly attenuated after adjusting for the referral to supervised exercise therapy. Further studies are needed to investigate whether the use of strategies to improve physical activity levels in US patients with PAD can improve their care and outcomes.

Table 3.

Between visit change in the proportion of sedentary patients with claudication among 1 908 patients with claudication at baseline, and 3, 6, and 12 months after enrolment in the Patient-centered Outcomes Related to Treatment Practices in Peripheral Arterial Disease: Investigating Trajectories (PORTRAIT) registry in the United States (US) and Netherlands between June 2011 and December 2015.

	US (n = 743)		Netherlands (n = 355)		Overall
Time	Change in sedentary patients — %	p value	Change in sedentary patients — %	p value	Change in sedentary patients — %	p value
Baseline-3 mo	−16.9	< .001	−4.2	.17	−12.7	< .001
Baseline-6 mo	−12.5	< .001	−10	< .001	−11.8	< .001
Baseline-12 mo	−7.7	.013	−13.2	< .001	−9.5	< .001
3 mo-6 mo	+4.4	.10	−5.8	< .001	+0.9	.62
3 mo-12 mo	+9.2	.020	−9	< .001	+3.2	.30
6 mo-12 mo	+4.8	.080	−3.2	< .001	+2.3	.26

WHAT THIS PAPER ADDS.

This is first study to examine patterns of physical inactivity in patients evaluated for symptomatic PAD between countries, and to provide insight into potential explanatory factors for observed differences in activity between the examined countries.