Decreased movement ability, one of the hallmarks of Parkinson disease (PD), may lead to inadequate physical activity (PA) and excessive time spent in sedentary behaviors—2 factors associated with an elevated risk for lifestyle-related diseases, poor management of PD, and premature death. To identify the extent to which people with PD are physically active, a comprehensive characterization of PA in this population is needed.
The study objective was to describe levels and patterns of PA and sedentary behaviors in elderly people with PD.
This cross-sectional study involved a free-living setting and 53 men and 42 women (mean age=73.4 years) with mild to moderate idiopathic PD.
Time spent in PA and sedentary behaviors was assessed for 1 week with accelerometers.
Mean daily step counts were 4,765; participants spent 589 minutes in sedentary behaviors, 141 minutes in low-intensity activities, 30 minutes in moderate-intensity lifestyle activities, and 16 minutes in moderate- to vigorous-intensity ambulatory activities. No differences were found between weekdays and weekend days. Patterns were characterized by a rise in total PA in the morning, peaking between 10 am and 3 pm, and a gradual decline toward the late evening. The proportion achieving 150 minutes of moderate- to vigorous-intensity PA per week was 27%, and 16% achieved 7,000 or more steps per day.
Nonrandomized selection of participants may limit the generalizability of the results.
Physical activity levels were generally low, in terms of both total volume and intensity, with only minor variations over the course of a day or between days. These results emphasize the need to develop strategies to increase PA and reduce time spent in sedentary behaviors in elderly people with mild to moderate PD.
Parkinson disease (PD) is a progressive, neurodegenerative disorder characterized by rigidity, tremor, impaired postural stability, decreased walking ability, and an increased risk of falls.1 These symptoms affect movement abilities in everyday tasks,1,2 leading to decreased physical activity (PA) and increased time spent in sedentary behaviors. This situation may, in turn, lead to a cycle of poorer balance, muscle weakness, fear of falling, and a sedentary lifestyle.
Insufficient PA is associated with many adverse health outcomes, and several studies have demonstrated PA to be effective in the treatment and prevention of a number of diseases.3–6 A systematic review recently highlighted the therapeutic value of PA,7 which reduces mortality for many common health conditions at rates similar to those of pharmaceutical treatments. In addition, PA is essential for healthy aging and has been shown, when in the form of structured exercise, to have positive effects on PD symptoms and complaints—such as improved gait performance, balance control, muscular strength, cardiovascular fitness, and quality of life (for reviews, see Speelman et al8 and Goodwin et al9).
Recent research based on self-report10 or accelerometry11 indicated that people with PD are approximately 30% less physically active than people in a control group matched for age. In addition, the volume of PA diminishes with a higher level of disease severity, as shown in both cross-sectional11 and longitudinal12 studies. However, little is known about any PA occurring in the lower range of the intensity spectrum. Emerging evidence has suggested that low-intensity PA (LPA) not only may decrease sedentary time—which, by itself, is a risk factor for disease5,6,13—but also may provide significant health benefits on its own.13 It was recently suggested that total daily activity counts may be more strongly associated with adverse health outcomes than moderate- to vigorous-intensity PA (MVPA)14 and, therefore, should be estimated along with time spent at different intensity levels in studies of PA (for a review, see Bassett et al15). To our knowledge, such data have not been presented for people with PD.
Another aspect of PA that is particularly relevant for PD is the diurnal pattern. Specifically, motor performance in people with PD has been shown to fluctuate over the course of a day, characterized primarily by morning activity levels that are lower than those of people who are healthy.16 However, to our knowledge, a more detailed characterization of diurnal variations, in which different intensity levels are quantified separately, has not been published.
Thus, the 3 aims of this study were: (1) to quantify the total volume of daily and weekly PA in elderly people with mild to moderate PD, (2) to determine the amount of time spent daily at different intensity levels, and (3) to characterize patterns of different intensity levels of PA and sedentary behaviors over the course of a day, including a comparison of weekdays with weekend days.
Method
Study Design
The present study was a substudy of an ongoing randomized controlled trial examining the effects of a balance training program on outcomes related to PD (the BETA-PD study; clinical trial number NCT01417598). The BETA study procedure has been described elsewhere.17 Accelerometer data collected at baseline (in 2012 and 2013) were analyzed in the present study.
Participants
People who had mild to moderate PD and dwelled in the community were recruited through advertisements in local newspapers, from Karolinska University Hospital, or from the Swedish Parkinson Association. Inclusion criteria were an age of 60 years or older and a clinical diagnosis of idiopathic PD, according to the Queen Square Brain Bank Criteria18 and Hoehn and Yahr stages 2 and 3.19 Exclusion criteria were atypical PD, according to Hughes et al20; a Mini-Mental State Examination score of ≤2421; and other neuromuscular disorders that influence gait and balance.
The study included 95 people (42 women) with idiopathic PD, classified as stage 2 (n=41) or 3 (n=54), according to Hoehn and Yahr.19 The means for various characteristics were as follows: age, 73.4 years (SD=5.7 years); weight, 76.6 kg (SD=14.2 kg); height, 171.4 cm (SD=9.3 cm); body mass index, 25.8 kg/m2 (SD=3.7); and disease duration, 5.9 years (SD=5.0 years). All participants provided written informed consent.
Instrumentation
Physical activity was assessed by use of ActiGraph GT3X+ accelerometers (ActiGraph, Pensacola, Florida) with firmware version 2.5.0. GT3X+ accelerometers sample changes in force at a frequency of 30 Hz before converting them to digital counts with a 12-bit analog-to-digital converter.22 To exclude nonhuman motion, accumulated data can be processed through 1 of 2 band-pass filters: either a normal filter (0.25–2.4 Hz) or a low-frequency extension option (bandwidth not specified by manufacturer). ActiGraph suggests applying the low-frequency extension filter for the assessment of people who exhibit very low acceleration outputs, such as elderly people.23 However, we have found this option likely to overestimate PA in people with PD and, therefore, selected the normal filter setting.24 One-minute epochs were used to determine the time spent at different intensity levels (definitions of PA levels are given below).
Data Acquisition
Upon arrival at the laboratory, participants were assessed for height and weight and evaluated with the motor part of the Unified Parkinson Disease Rating Scale.25 Each participant was then provided with an accelerometer attached to a belt and given verbal and written instructions on its proper use. The belt was to be worn around the waist at hip level during all time awake for 7 consecutive days, except when showering, swimming, or bathing.
Data Reduction
ActiLife 6 (ActiGraph) was used to process all accelerometer data: data cleansing, filtering, and computation of different PA parameters. In accordance with previous research on elderly people, episodes ≥90 minutes of consecutive zeroes (ie, inactivity) were excluded.26 Wear time was determined by subtracting nonwear time from total time. Hours with ≥54 minutes of data (90%) were included.
The following outcome variables were generated: total volume of PA, defined as vector magnitude counts per minute; steps per day; and minutes spent at different intensity levels per day, based on vertical axis counts. Total daily vector magnitude counts, representing the sum of averaged triaxial vector counts per minute, have been suggested to provide a more reliable means of operationalizing PA than estimating time at different intensity levels15 and, therefore, also were included. Because the ActiGraph system has not been calibrated for people with PD and preliminary analyses demonstrated zero or very few minutes at vigorous intensity per week—that is, greater than 6 metabolic equivalents (METs)—cutoffs that were previously used in epidemiological studies of elderly people27,28 were applied as follows: for sedentary behaviors, 1 to 1.5 METs or 0 to 99 cpm; for LPA, 1.5 to 2.9 METs or 100 to 759 cpm; and for moderate-intensity PA or higher (MPA+), ≥3 METs or ≥760 cpm. Moderate-intensity PA or higher was further divided into moderate-intensity lifestyle activities (MPALS; 3 METs or 760–2,019 cpm), for activities such as sweeping, vacuuming, raking, and shoveling,27 and moderate- to vigorous-intensity ambulatory activities (MVPA; >3 METs or >2,019 cpm), corresponding to a walking speed of 4 km/h or more.28
Data Analysis
Levels
In accordance with previous research,28–30 days with ≥10 hours of valid data were used to summarize the total amount of daily PA and the time spent at different intensity levels in terms of means and 95% confidence intervals (CIs). New variables representing average time at different intensity levels per day (Monday–Sunday) were determined from the mean values for a minimum of 4 and up to 7 days per week, depending on how many valid days were available. The requirement for 4 recording days was based on previous research on adults who were healthy31 and adults who were overweight,32 and the assessment of 4 days of PA was shown to agree highly with the assessment of a full week of PA.29,31
To determine the proportion meeting the guideline of 150 minutes of MVPA weekly or 7,000 or 8,000 steps per day, proportions and 95% CIs (binomial exact method) were estimated. Average values for 4 to 7 valid days were multiplied by 7 to obtain time spent in MVPA per week. The cutoff for meeting the PA recommendation was set to ≥150 minutes of accumulated time spent in MVPA per week. Two cutoffs for meeting the recommendation for total daily steps were used: 7,000 steps and 8,000 steps.
Patterns
Outcome variables representing weekdays were generated from the mean values for 3 to 5 weekdays (ie, Monday–Friday), and outcome variables representing weekend days were determined from the mean values for both Saturday and Sunday. That is, a minimum of 3 weekdays or both weekend days, respectively, was required for inclusion in the analyses. Statistical analyses were performed with STATA version 11 (StataCorp LP, College Station, Texas). All variables demonstrated an approximately normal distribution when evaluated on a histogram, except for MVPA, which was skewed to the right. Log10 and natural log transformations did not improve the distribution; hence, nontransformed data were used. Differences between weekdays and weekends with respect to time spent at different PA intensity levels and steps per day were estimated with a one-way repeated-measures analysis of variance or a Wilcoxon signed rank test for MVPA. The alpha level was set at P≤.05.
To describe the patterns of PA and sedentary behaviors over the course of a day, we plotted time (minutes per hour) spent at different intensity levels over an 18-hour time frame between 6 am and 11 pm for weekdays and weekend days. Because the values for MVPA and MPALS were very low and difficult to evaluate on a graph, their combined values (MPA+) were used to illustrate the number of minutes spent in PA of moderate or higher intensity each hour. All valid hours within the given time frame were selected to calculate the average number of minutes spent at different intensity levels for each time point (ie, for each hour).
Role of the Funding Source
The study was funded by the Swedish Research Council, StratNeuro Karolinska Institutet, Norrbacka-Eugenia Foundation, and the national doctoral school in health care sciences at Karolinska Institutet.
Results
Participant characteristics are shown in Table 1. Valid data were obtained from 79 participants for a minimum of 4 of 7 days per week, from 79 participants for 3 to 5 weekdays, from 62 participants for weekend days, and from 59 participants for both weekdays and weekend days.
Characteristics of Participants From Whom Valid Data Were Obtained for Various Periodsa
| Characteristic | Minimum of 4/7 d/wk (n=79) | 3–5 Weekdays (n=79) | Weekend Daysb (n=62) |
|---|---|---|---|
| Age, y | 73.2 (5.7) | 73.3 (5.7) | 73.5 (5.9) |
| Women, n (%) | 37 (47) | 36 (46) | 29 (47) |
| Height, cm | 171.5 (8.8) | 171.8 (9.1) | 171.9 (8.4) |
| Weight, kg | 75.1 (13.9) | 75.6 (14.4) | 75.9 (12.8) |
| BMI, kg·m2 | 25.4 (3.7) | 25.5 (3.7) | 25.6 (3.6) |
| UPDRS score, mean (range) | 35.3 (14–62) | 35.7 (14–62) | 34.8 (17–62) |
| Characteristic | Minimum of 4/7 d/wk (n=79) | 3–5 Weekdays (n=79) | Weekend Daysb (n=62) |
|---|---|---|---|
| Age, y | 73.2 (5.7) | 73.3 (5.7) | 73.5 (5.9) |
| Women, n (%) | 37 (47) | 36 (46) | 29 (47) |
| Height, cm | 171.5 (8.8) | 171.8 (9.1) | 171.9 (8.4) |
| Weight, kg | 75.1 (13.9) | 75.6 (14.4) | 75.9 (12.8) |
| BMI, kg·m2 | 25.4 (3.7) | 25.5 (3.7) | 25.6 (3.6) |
| UPDRS score, mean (range) | 35.3 (14–62) | 35.7 (14–62) | 34.8 (17–62) |
Data are presented as means (standard deviations) unless indicated otherwise. BMI=body mass index, UPDRS=Unified Parkinson Disease Rating Scale (motor part).
Saturday and Sunday.
Characteristics of Participants From Whom Valid Data Were Obtained for Various Periodsa
| Characteristic | Minimum of 4/7 d/wk (n=79) | 3–5 Weekdays (n=79) | Weekend Daysb (n=62) |
|---|---|---|---|
| Age, y | 73.2 (5.7) | 73.3 (5.7) | 73.5 (5.9) |
| Women, n (%) | 37 (47) | 36 (46) | 29 (47) |
| Height, cm | 171.5 (8.8) | 171.8 (9.1) | 171.9 (8.4) |
| Weight, kg | 75.1 (13.9) | 75.6 (14.4) | 75.9 (12.8) |
| BMI, kg·m2 | 25.4 (3.7) | 25.5 (3.7) | 25.6 (3.6) |
| UPDRS score, mean (range) | 35.3 (14–62) | 35.7 (14–62) | 34.8 (17–62) |
| Characteristic | Minimum of 4/7 d/wk (n=79) | 3–5 Weekdays (n=79) | Weekend Daysb (n=62) |
|---|---|---|---|
| Age, y | 73.2 (5.7) | 73.3 (5.7) | 73.5 (5.9) |
| Women, n (%) | 37 (47) | 36 (46) | 29 (47) |
| Height, cm | 171.5 (8.8) | 171.8 (9.1) | 171.9 (8.4) |
| Weight, kg | 75.1 (13.9) | 75.6 (14.4) | 75.9 (12.8) |
| BMI, kg·m2 | 25.4 (3.7) | 25.5 (3.7) | 25.6 (3.6) |
| UPDRS score, mean (range) | 35.3 (14–62) | 35.7 (14–62) | 34.8 (17–62) |
Data are presented as means (standard deviations) unless indicated otherwise. BMI=body mass index, UPDRS=Unified Parkinson Disease Rating Scale (motor part).
Saturday and Sunday.
Levels of PA
Of the 13 hours 6 minutes of wear time for an average day of the week, 9 hours 49 minutes (75%) was spent in sedentary behaviors, 2 hours 21 minutes (18%) was spent in LPA, 30 minutes (4%) was spent in MPALS, and 16 minutes (2%) was spent in MVPA. The mean total time spent in MVPA per week was 114.8 minutes (SD=150.4 minutes). Table 2 shows means and 95% CIs for total time spent in PA (vector magnitude counts per day and per minute and step counts per day) as well as time spent at different PA intensity levels and in sedentary behaviors for an average day of the week, weekday, and weekend day.
Accelerometer Outcomes Determined for Various Periodsa
| Activity | Minimum of 4/7 d/wk (n=79) | 3–5 Weekdays (n=79) | Weekend Daysb (n=62) |
|---|---|---|---|
| SB, min·d−1 | 588.9 (571.6, 606.1) | 593.9 (575.6, 612.3) | 584.5 (561.0, 608.0) |
| LPA, min·d−1 | 140.6 (126.5, 154.6) | 143.0 (128.7, 157.2) | 142.1 (124.2, 160.0) |
| MPALS, min·d−1 | 30.1 (25.3, 34.9) | 30.7 (26.0, 35.4) | 31.3 (23.6, 39.0) |
| MVPA, min·d−1 | 16.4 (11.6, 21.2) | 15.2 (10.8, 19.7) | 18.4 (11.1, 25.7) |
| Steps per day | 4,765 (4,071, 5,461) | 4,721 (4,038, 5,403) | 4,888 (3,896, 5,881) |
| VM, counts·d−1 | 293,614 (259,255, 327,972) | 291,470 (256,899, 326,041) | 305,269 (262,068, 348,469) |
| VM, counts·min−1 | 374 (77, 1,146) | 368 (73, 1,144) | 388 (83, 1,229) |
| Wear time, h·d−1 | 13.1 (12.7, 13.4) | 13.2 (12.8, 13.6) | 13.1 (12.7, 13.6) |
| Activity | Minimum of 4/7 d/wk (n=79) | 3–5 Weekdays (n=79) | Weekend Daysb (n=62) |
|---|---|---|---|
| SB, min·d−1 | 588.9 (571.6, 606.1) | 593.9 (575.6, 612.3) | 584.5 (561.0, 608.0) |
| LPA, min·d−1 | 140.6 (126.5, 154.6) | 143.0 (128.7, 157.2) | 142.1 (124.2, 160.0) |
| MPALS, min·d−1 | 30.1 (25.3, 34.9) | 30.7 (26.0, 35.4) | 31.3 (23.6, 39.0) |
| MVPA, min·d−1 | 16.4 (11.6, 21.2) | 15.2 (10.8, 19.7) | 18.4 (11.1, 25.7) |
| Steps per day | 4,765 (4,071, 5,461) | 4,721 (4,038, 5,403) | 4,888 (3,896, 5,881) |
| VM, counts·d−1 | 293,614 (259,255, 327,972) | 291,470 (256,899, 326,041) | 305,269 (262,068, 348,469) |
| VM, counts·min−1 | 374 (77, 1,146) | 368 (73, 1,144) | 388 (83, 1,229) |
| Wear time, h·d−1 | 13.1 (12.7, 13.4) | 13.2 (12.8, 13.6) | 13.1 (12.7, 13.6) |
Data are presented as means (95% confidence intervals). SB=sedentary behaviors, LPA=low-intensity physical activities, MPALS=moderate-intensity lifestyle activities, MVPA=moderate- to vigorous-intensity physical activity, VM=vector magnitude. No statistically significant differences were found between weekdays and weekend days for any of the outcomes (P>.05).
Saturday and Sunday.
Accelerometer Outcomes Determined for Various Periodsa
| Activity | Minimum of 4/7 d/wk (n=79) | 3–5 Weekdays (n=79) | Weekend Daysb (n=62) |
|---|---|---|---|
| SB, min·d−1 | 588.9 (571.6, 606.1) | 593.9 (575.6, 612.3) | 584.5 (561.0, 608.0) |
| LPA, min·d−1 | 140.6 (126.5, 154.6) | 143.0 (128.7, 157.2) | 142.1 (124.2, 160.0) |
| MPALS, min·d−1 | 30.1 (25.3, 34.9) | 30.7 (26.0, 35.4) | 31.3 (23.6, 39.0) |
| MVPA, min·d−1 | 16.4 (11.6, 21.2) | 15.2 (10.8, 19.7) | 18.4 (11.1, 25.7) |
| Steps per day | 4,765 (4,071, 5,461) | 4,721 (4,038, 5,403) | 4,888 (3,896, 5,881) |
| VM, counts·d−1 | 293,614 (259,255, 327,972) | 291,470 (256,899, 326,041) | 305,269 (262,068, 348,469) |
| VM, counts·min−1 | 374 (77, 1,146) | 368 (73, 1,144) | 388 (83, 1,229) |
| Wear time, h·d−1 | 13.1 (12.7, 13.4) | 13.2 (12.8, 13.6) | 13.1 (12.7, 13.6) |
| Activity | Minimum of 4/7 d/wk (n=79) | 3–5 Weekdays (n=79) | Weekend Daysb (n=62) |
|---|---|---|---|
| SB, min·d−1 | 588.9 (571.6, 606.1) | 593.9 (575.6, 612.3) | 584.5 (561.0, 608.0) |
| LPA, min·d−1 | 140.6 (126.5, 154.6) | 143.0 (128.7, 157.2) | 142.1 (124.2, 160.0) |
| MPALS, min·d−1 | 30.1 (25.3, 34.9) | 30.7 (26.0, 35.4) | 31.3 (23.6, 39.0) |
| MVPA, min·d−1 | 16.4 (11.6, 21.2) | 15.2 (10.8, 19.7) | 18.4 (11.1, 25.7) |
| Steps per day | 4,765 (4,071, 5,461) | 4,721 (4,038, 5,403) | 4,888 (3,896, 5,881) |
| VM, counts·d−1 | 293,614 (259,255, 327,972) | 291,470 (256,899, 326,041) | 305,269 (262,068, 348,469) |
| VM, counts·min−1 | 374 (77, 1,146) | 368 (73, 1,144) | 388 (83, 1,229) |
| Wear time, h·d−1 | 13.1 (12.7, 13.4) | 13.2 (12.8, 13.6) | 13.1 (12.7, 13.6) |
Data are presented as means (95% confidence intervals). SB=sedentary behaviors, LPA=low-intensity physical activities, MPALS=moderate-intensity lifestyle activities, MVPA=moderate- to vigorous-intensity physical activity, VM=vector magnitude. No statistically significant differences were found between weekdays and weekend days for any of the outcomes (P>.05).
Saturday and Sunday.
Estimations of binomial proportions demonstrated that 27% (95% CI=18%, 36%) met the guideline of 150 minutes of MVPA per week. The proportion achieving 7,000 or more steps per day was 16% (95% CI=8%, 25%), whereas 11% (95% CI=4%, 19%) achieved 8,000 or more steps per day.
Patterns of PA
There were nonsignificant trends toward more time being spent in sedentary behaviors (mean difference=9 minutes) and LPA (mean difference=0.9 minute) on weekdays and more time being spent in higher-intensity PA on weekend days (mean difference in MPALS=0.6 minute; mean difference in MVPA=3.2 minutes; mean difference in steps=167) (Tab. 2). Similarly, vector magnitude counts per day (mean difference=13,799) and per minute (mean difference=167) were higher on weekend days than on weekdays; however, these differences were not statistically significant (all P values >.5). The differences were even smaller when wear time (5 minutes longer on weekdays) was taken into account; when expressed as a percentage of daily wear time, sedentary behaviors represented 75% of weekdays and 74% of weekend days, and LPA, MPALS, and MVPA represented 18%, 4%, and 2%, respectively, of both weekdays and weekend days (Tab. 2). Descriptive data for weekdays and weekend days are shown in Table 2.
The Figure shows PA patterns over the course of a day, with weekdays and weekend days being displayed separately. Physical activity patterns over the course of a day showed that LPA increased starting at approximately 7 am, climbed toward a peak at 10 to 11 am, and then gradually tapered off toward a low plateau at approximately 9 pm. A slight increase was again observed between 10 and 11 pm. Weekdays and weekend days showed virtually identical patterns at this intensity level. Time in MPA+ followed a similar pattern, although at a consistently lower level. Again, similar patterns were found between weekdays and weekend days, with 2 exceptions: the first between 1 and 3 pm and the second between 10 and 11 pm. At both of these times, MPA+ tended to be higher on weekend days than on weekdays. This elevation was paralleled by a decrease in sedentary behaviors during the same times. Apart from this difference, patterns of sedentary behaviors were similar on weekdays and weekend days.
Illustration of the average number of minutes spent in sedentary behaviors (prefix S), low-intensity physical activities (prefix L), and a combination of moderate-intensity lifestyle activities and moderate- to vigorous-intensity ambulatory activities (prefix M+) on an hour-to-hour basis for weekdays (W; black symbols) and weekend days (WE; gray symbols).
Discussion
The main finding of the present study was that approximately 75% of all time awake was spent in sedentary behaviors, 18% was spent in LPA, and 6% was spent in a combination of MPALS or MVPA. These results are in line with previous research demonstrating that people with mild to moderate PD are ambulatory at moderate or higher intensity only 5% of the day.11 They also largely agree with the previous finding that 76% of all time awake comprises sedentary behaviors.33 When explicitly studying people who had PD and led a sedentary lifestyle, other authors found that 98% of all time awake comprised a combination of sedentary behaviors and LPA, whereas less than 2% represented moderate- to vigorous-intensity PA.34 Despite several methodological differences among these studies, the results collectively indicate that people with PD lead a lifestyle characterized by excessive sedentary behaviors and insufficient MPA+.
The findings of the present study have important clinical implications because both excessive time in sedentary behaviors and insufficient MVPA predispose people to many comorbidities, such as coronary heart disease, stroke, diabetes, hypertension, cancer, depression, and premature death.3–6 In addition, a sedentary lifestyle may have a direct negative effect on disease management and progression because the potentially positive effects of exercise on symptoms specifically related to PD are lost.8
Recent epidemiological studies suggested that LPA and sedentary behaviors are independently associated with metabolic syndrome13 and that total activity counts are more strongly associated with biomarkers for cardiovascular disease than bout-accumulated MVPA.14 On the basis of these findings and others, it has been proposed that total daily PA, in terms of activity counts, should be considered in the evaluation of PA, rather than time spent in MVPA alone.15 In the present study, we found that 18% of the day was spent in LPA, which together with MPALS and MVPA adds up to a little more than 3 hours in any PA daily and total daily activity counts to 293,614. Additional research that relates daily activity counts to adverse health outcomes is needed to interpret these results.
In a few studies, PA behavior in people with PD and PA behavior in people in a similarly aged control group were compared directly. Using questionnaire data, van Nimwegen et al10 found that people with PD were 29% less physically active than people in a control group. Similarly, using accelerometry, Lord et al11 found that the level of ambulatory PA was 27% lower in people with PD than in a control group of people who were healthy. Lord et al11 also found that the step count was 5,500 steps per day; this count was similar to the count of 4,760 steps per day reported here but in stark contrast to the daily count of 10,200 steps reported by Cavanaugh et al.12 This rather large discrepancy is most likely due to methodological differences, further emphasizing the need to standardize PA measurements across studies.
When we evaluated diurnal PA patterns, only minor fluctuations were seen over the course of a day, confirming previous research.16 Expanding on what was previously known, we found patterns of low variability across all intensity levels. On weekdays, PA levels peaked at about 10 to 11 am before gradually declining, whereas they remained elevated until the early afternoon on weekend days. This finding suggests that motor symptoms may be less prominent at midday and during the early afternoon, even though this notion was not explicitly studied. Although these findings provide a better understanding of PA patterns in people with PD on a group level, several different mechanisms—including fatigue, fear of falling, lack of motivation, physical constraints, and symptoms related to pharmacological treatment—could underpin PA on an individual level. Therefore, to provide each patient with the most appropriate recommendations, PA programs may need to be individually tailored on the basis of a thorough anamnesis and accelerometer assessment of each patient.
In addition to describing the levels and patterns of PA, we also estimated the extent to which PA recommendations were met in the studied population; we found that 73% of the participants failed to achieve 150 minutes of MVPA per week. Most studies evaluating PA in various populations use the guidelines set forth by the American College of Sports Medicine (2008),4 according to which older adults are advised to perform a total of 150 minutes of MVPA per week, preferably in bouts no shorter than 10 minutes. These guidelines were derived from self-reported data, which have been shown to agree rather poorly with objective methods, generally returning higher levels of PA than accelerometry.35 This incongruence becomes particularly striking when the rule of 10-minute bouts of accumulated MVPA is applied; whereas one study based on self-report estimated that more than 46% of the general adult population achieved 150 minutes of MVPA (in bouts) per week,36 a study based on accelerometry found that less than 5% reached the same level.28 Discrepancies between self-reported PA and accelerometry also have been found in studies of people with PD specifically.37 Collectively, these studies demonstrate that PA recommendations derived from self-reports likely do not apply to accelerometry.
Taken together, our results demonstrate that people with mild to moderate PD spend a large part of their day in sedentary behaviors and only a small amount of time in PA. These findings are important to physical therapists and other health care professionals who may promote and prescribe PA. Although the benefits of PA in disease prevention and management are now widely recognized, the specific methods and procedures that may be effective for increasing PA in people with mild to moderate PD have yet to be determined.
Limitations
One of the main challenges for assessing PA by means of accelerometry is to define appropriate cutoffs to represent different types of PA. Because the ActiGraph system has not been calibrated for people with PD, we used cutoffs that have been developed from the general population. It is possible that these thresholds do not validly reflect intensity levels in people with PD because, among other issues, PD generally presents with an altered gait pattern that is likely more energy demanding than “normal” gait.38
The present study was part of an ongoing project in which participants were recruited locally rather than being randomly selected from the whole population of people with PD. The way in which this method may have affected the results is not easily discerned; however, it likely limits the generalizability of the study. Moreover, we specifically included people with mild to moderate PD, and additional studies are needed to characterize PA and sedentary behaviors in people with more severe PD.
The present study demonstrated that elderly people with mild to moderate PD displayed low levels of PA, in terms of both total volume and intensity. Much of their day was spent in sedentary behaviors, with small variations in PA over the course of a day and over different days of the week. Taken together, the results emphasize the need to develop strategies that will increase PA levels and reduce time spent in sedentary behaviors in this population.
The Regional Board of Ethics in Stockholm, Sweden, granted approval for the study.
The study was funded by the Swedish Research Council, StratNeuro Karolinska Institutet, Norrbacka-Eugenia Foundation, and the national doctoral school in health care sciences at Karolinska Institutet.
References
Author notes
Dr Benka Wallén, Dr Franzén, and Dr Hagströmer provided concept/idea/research design. All authors provided writing. Dr Benka Wallén, Dr Franzén, and Mr Nero provided data collection. Dr Benka Wallén and Dr Hagströmer provided data analysis. Dr Franzén, Mr Nero, and Dr Hagströmer provided project management. Dr Franzén and Dr Hagströmer provided fund procurement, participants, and facilities/equipment. Dr Franzén provided institutional liaisons. Mr Nero provided administrative support and consultation (including review of manuscript before submission). The authors acknowledge PhD students David Conradsson and Niklas Löfgren for support with data collection.
Comments