This study hypothesized that IC patients would demonstrate a lower PA level and would comply less frequently to the lower limit of the ACSM/AHA recommendations for PA and public health compared to healthy adults. In a two-day assessment of PA using a vertical accelerometer and a pedometer, one other study found that daily PA level of IC patients was 42%-45% lower compared to healthy subjects of similar age.17 Also using a vertical accelerometer, a second study demonstrated that patients with peripheral arterial occlusive disease (PAOD) had markedly lower PA levels compared to non-PAOD participants (803 ± 364 kcal/wk vs. 1750 ± 1296 ± kcal/wk).15 Ever since the ACSM/AHA issued minimum recommendations on PA and public health were issued, studies describing to what extent IC patients meet these criteria in relation to healthy adults were not performed.

 

 The present study using newest generation of tri-axial accelerometers found that fewer than half of IC patients meet the ACSM/AHA minimum recommendations while more than one-quarter of older healthy adults do not meet these criteria.

Moreover, the daily level of PA of IC patients (387 ± 198) is much lower compared to healthy adults (500 ± 156). IC patients reporting substantial functional impairment (WIQ score <0.4) had the lowest mean daily PA level of all groups. Interestingly, the PA group difference was solely attributed to this subgroup. Only minor functional impairment in IC patients apparently does not result in lower daily PA level.

 

 A tri-axial accelerometer allows for separately identifying various activities such as sitting, standing, lying or walking. The present study found significant group differences regarding time of sitting and 'not wearing the device'. Healthy adults appearly sit approximately one hour longer and might sleep less compared to IC patients. In contrast, walking time did not differ among groups, although the daily number of steps and highest number of steps were significantly lower in the IC group. This difference is possibly explained by a slower walk pace or cadence associated with IC. A previous study also suggested that patients with IC take fewer steps and have a lower daily average cadence compared to healthy adults.32

 

 Variables influencing daily PA level in IC patients were hitherto not identified.

The present study revealed that age, previous history of pulmonary disease, MWD and the WIQ, SF-12 MCS and SF-12 PCS score all independently influence these levels. Pulmonary history appeared the strongest factor. However, patient with serious pulmonary conditions (affecting the patient's walking ability) were excluded from this study. Therefore, the effect of pulmonary history on the daily PA level in an IC patient is probably underestimated. Patients with COPD are markedly inactive in daily life
compared to healthy adults.33 Therefore, physicians need to realize that patients with IC and COPD are much less active compared to IC patients without pulmonary history or healthy adults.

 

 A low daily PA level is generally considered a strong predictor of mortality and functional decline in patients with IC.6-9 More than light-intensity weekly PA improves 5-year survival rate significantly compared to light-intensity or no physical activity.7

For this reason, one could argue that the PA level is a valid outcome parameter in evaluating treatment effects of any intervention for IC. Current outcome parameters including maximum- or pain-free walking distances only describe the patient's exercise capacity while it does not reflect to what extent this capacity is used. In contrast, daily PA levels, provide objective information on the frequency, duration and intensity of exercise. Combined with a functional impairment score such as the WIQ-score, physicians may objectively monitor the patient's treatment progress.
Future studies should focus on the effects of different treatment strategies on PA. A rehabilitation program such as supervised exercise therapy (SET) improves walking performance and functional impairment. Most programmes promote, besides supervised sessions, daily home-based exercise in order to increase the patient's daily PA. No studies have substantiated the effect of SET on the daily level of PA yet. Therefore, accelerometer studies before and after different treatment strategies may shed light on these issues.

Study limitations

 The present study is limited by the number of excluded participants. In total, 38 participants in the IC group (40.4 %) and 9 controls (25.0 %) were excluded because the device was not correctly worn for =5 days. By definition, a correctly worn day was considered a day that the device was worn for =10 hours, not worn for =5 hours and with at least 20 steps. These strict criteria were a priori determined to prevent bias. For instance, if a participant wore the device for only 5 hours one day, unrepresentative data on the overall person's PA would have been collected. The high dropout rate may have potentially biased the PA results, most likely an overestimation of the current PA level in IC patients. Lack of motivation or misunderstanding the monitoring protocol could have led to incorrect use. A substantial number of participants said they had forgotten the end date of the monitoring period. In a few instances, participants fell asleep with the device still strapped to their waist. In our monitoring protocol, participants were asked to remove the device before sleeping to improve compliance. If the device was worn during night, that day was considered by definition 'not correctly worn'. We are not sure if compliance could have been higher if participants wore the device constantly. In future studies, regular check-ups by phone calls could be relevant as a means to increase compliance and understanding of the monitoring protocol. For clinical practice one could consider a shorter amount of the monitoring period, considering results are summarized in an average daily PA level. However, accuracy of PA results of a shorter monitoring period should first be validated since these could vary.

 

 We experienced only a few measurement failures with the tri-axial

accelerometer. The Dynaport MoveMonitor can fail if the battery is not fully charged or if the device is worn backwards. The device indicates if it is fully charged, minimizing the change of human errors. However, in a handful of cases the device did not measure for 7 days because the investigators replaced the original SD-cards with new ones. The new SD-cards appeared to consume more battery energy than the original ones. Therefore the device could only measure for 5 or 6 days. Besides this error, we did not experience device malfunction. The investigators explicitly explained how to wear the device to minimize the change that participants wore the device backwards. This could be explained quite easy since the device is worn correct if one could read the word 'McRoberts' on the outside of the device. Therefore this error did not occur during any of the worn days.
Another study limitation is the lack of an age-matched control group. Although groups were initially matched for age and sex, a significant difference in age was present between the IC and control group once the IC group drop out rate appeared high. However, a sensitivity analysis including healthy adults aged = 60 years (n=14) revealed no significant difference in daily PA level in comparison to 13 other healthy adults aged > 60 years (526 ± 160 METs?min vs. 471 ± 151 METs?min, p=0.37) although these groups may have been underpowered.