Sage Journals: Discover world-class research

Abstract

Introduction. Increasingly physical activity (PA) is being recognized as an important outcome in non–small cell lung cancer (NSCLC). We investigated PA using novel global positioning system (GPS) tracking individuals with NSCLC and a group of similar-aged healthy individuals. Methods. A prospective cross-sectional multicenter study. Fifty individuals with NSCLC from 3 Australian tertiary hospitals and 35 similar-aged healthy individuals without cancer were included. Individuals with NSCLC were assessed pretreatment. Primary measures were triaxial accelerometery (steps/day) and GPS tracking (outdoor PA behavior). Secondary measures were questionnaires assessing depression, motivation to exercise, and environmental barriers to PA. Between-group comparisons were analyzed using analysis of covariance. Results. Individuals with NSCLC engaged in significantly less PA than similar-aged healthy individuals (mean difference 2363 steps/day, P = .007) and had higher levels of depression (P = .027) and lower motivation to exercise (P = .001). Daily outdoor walking time (P = .874) and distance travelled away from home (P = .883) were not different between groups. Individuals with NSCLC spent less time outdoors in their local neighborhood area (P < .001). A greater number of steps per day was seen in patients who were less depressed (r = .39) or had better access to nonresidential destinations such as shopping centers (r = .25). Conclusion. Global positioning system tracking appears to be a feasible methodology for adult cancer patients and holds promise for use in future studies investigating PA and or lifestyle behaviors.

Keywords

physical activity lung cancer global positioning system (GPS) tracking outdoor behavior exercise

Introduction

Non–small cell lung cancer (NSCLC) is associated with significant disease burden, deterioration in physical function, and impairment in health-related quality of life.^1-4 Individuals with NSCLC commonly experience debilitating symptoms, which often include a combination of breathlessness, fatigue, cough, and pain. After a diagnosis of NSCLC, a cycle of functional decline and inactivity commonly ensues. The American College of Sports Medicine recommends that adults with cancer engage in 30 minutes of moderate-intensity physical activity (PA) on 5 or more days of the week.⁵ These are the same guidelines as those that apply to the general older adult population.⁶ However, in comparison to other types of cancer, there has been a lack of research quantifying PA behavior of people with NSCLC, particularly with the use of objective measurement tools.

Previous research has shown that time spent outdoors is associated with improved physical and mental health outcomes,^7,8 including lower levels of depression and higher self-esteem in the general population.⁹ Global positioning system (GPS) tracking provides a novel and unique method of measuring time spent outdoors objectively; however, no studies have previously used this in a cancer population. The primary aims of this study were, therefore, to use GPS tracking to (1) objectively measure level of PA (both total and outdoors) and (2) assess total time spent outdoors by individuals with newly diagnosed NSCLC and compare this to similar-aged healthy individuals. Secondary aims were to (3) compare motivation to exercise between individuals with NSCLC and similar-aged healthy individuals and to (4) explore the relationships between PA and (a) levels of depression, (b) motivation to exercise, and (c) environmental barriers to PA.

Methods

A prospective cross-sectional multicenter study was conducted from November 2010 to October 2012. Two groups of participants aged 18 years or older were studied. Group 1 consisted of individuals with newly diagnosed nonmetastatic (stage I to IIIB) NSCLC (Table 1) who were screened and recruited from weekly outpatient clinics by recruitment therapists at 3 tertiary hospitals in Melbourne, Australia. Group 2 consisted of stable community-dwelling individuals without cancer (Table 1). Healthy individuals were recruited from posters advertising the study. All sites had institutional ethical approval, and participants provided written informed consent. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were followed to report this study.^10,11

Table 1.

Inclusion and Exclusion Criteria.

Criteria	Group 1: Lung Cancer	Group 2: Healthy Individuals
Inclusion	• Aged 18 years or older	• Aged 18 years or older
	• Histological confirmed stage I to IIIB non–small cell lung cancer	• No history or current diagnosis of cancer (excluding individuals with previous prostate or nonmelanoma skin cancer, disease free for at least 10 years)
	• Prior to cancer treatment: surgery, chemotherapy, or radiotherapy
Exclusion	• Stage IV lung cancer	• Insufficient English language skills to complete questionnaires
	• Recurrent cancer disease	• Severe active psychiatric or cognitive disorder (chronic or acute)
	• Insufficient English language skills to complete questionnaires (decided at the discretion of the recruitment therapist)	• Inability to provide informed consent
	• Severe active psychiatric or cognitive disorder (chronic or acute)	• Musculoskeletal condition precluding assessment of walking (such as amputation)
	• Inability to provide informed consent
	• Musculoskeletal condition precluding assessment of walking (such as amputation)

Demographic data were obtained including living arrangements, employment status, limitation to walking, and use of gait aids. Participants living further than 50 km from Melbourne city were classified as living in a rural location. Performance status was rated by the participant and physician (NSCLC group only) using the Eastern Cooperative Oncology Group–Performance Status (ECOG-PS) scale, which ranges from 0 (fully active) to 4 (bed bound)/5 (deceased).¹² Comorbidities were scored with the simplified Colinet comorbidity score (higher scores represent greater comorbid disease)¹³ and respiratory function was collected. Medical data, for individuals with NSCLC, were collected including tumor stage, tumor histology, and type of cancer treatment (surgery, chemotherapy, and/or radiotherapy) undertaken in the subsequent first 6 months following diagnosis.

Physical activity was measured using a KinetaMap device (Sparkfun Electronics GPS-08725, Boulder, CO), which incorporated triaxial accelerometery (number of steps per day) and GPS tracking (time spent walking outdoors, time spent outdoors, and distance travelled from the home).¹⁴ Participants were instructed to wear the device around their waist for 5 consecutive days including a weekend day and were required to charge the device overnight. A minimum of 3 “full days” of data (defined as device turned “on” for ≥8 h/day) were required for participants’ data to be included in analyses. Data were analyzed with computer software programs custom-designed by our research personnel.¹⁵ All accelerometer and GPS data were averaged across the number of full days that the device were worn. The accelerometer component of the KinetaMap device registered time-stamped acceleration of the body in 3 planes (sampled at 50 Hz), and data were exported into the computer program to calculate steps per day. Raw GPS coordinates were sampled at 1 Hz and inputted into the computer program, which loaded data into Google Earth and plotted the coordinate and velocity related data on a graph. This allowed for the manual identification of periods of outdoor and indoor activity, with the locations of travel and velocity of movement used to determine whether the activity was likely to be walking or passive transport related. Outdoor time was also split into 4 distinct regions, namely, (1) at home (<100 m from home), (2) close to home (100 m to 1 km from home), (3) local neighborhood area (1-5 km from home), and (4) outside the neighborhood (>5 km from home). This geographic division of outdoor time allowed for the analysis of patterns of movement relative to the individual’s home, which may be affected in the NSCLC group due to factors such as their need to attend clinical appointments. Participants were asked to recall, over the previous 7 days, the number of sessions of PA undertaken, the type of most frequent PA performed (eg, walking, golf, cycling), and their average daily television viewing time (in hours).

Measurement of Secondary Outcomes

Levels of depression were measured using the Hospital Anxiety and Depression Scale (HADS).¹⁶ The HADS-depression scale is composed of 7 questions, scored on a 4-point Likert-type scale from 0 to 3, and asks participants to recall how they have been feeling over the past 7 days. Answers are summated to give a score out of 21, and higher scores represent higher levels of depression. Motivation for exercise was assessed using the 19-item Behavioral Regulation of Exercise Questionnaire Version 2 (BREQ-2),¹⁷ which has been validated in healthy¹⁸ and colorectal cancer populations.¹⁹ The 19-item questionnaire is composed of 5 subscales, and responses to the 19 statements are rated on a 5-point Likert-type scale from 0 (not true for me) to 4 (very true for me). The following definitions apply to the subscales: external regulation (“engaging in PA due to external pressures or to achieve externally imposed rewards”); introjected regulation (“internalization of external controls which are then applied through self-imposed pressures in order to avoid guilt or to maintain self-esteem”); identified regulation (“conscious acceptance of the PA as being important in order to achieve personally valued outcomes”); intrinsic regulation (“undertaking PA for enjoyment or satisfaction inherent in engaging in the actual PA”); and amotivation ("a state of lacking any intention to engage in a behavior”).¹⁷ The mean item scores for the 5 BREQ-2 subscales were calculated and an overall total Relative Autonomy Index was also calculated by applying published weightings to subscores and summating total scores.¹⁷ Higher regulation subscores represent more motivation to exercise. Higher amotivation scores represent less motivation to exercise. Higher Relative Autonomy Index scores represent more autonomous motivation and greater self-determination to exercise.

Environmental barriers to PA were assessed using the 65-item Neighborhood Environment Walkability Scale–Australian Version (NEWS-AU), a valid and reliable tool²⁰ designed to assess factors associated with the residential neighborhood that may have an impact on an individual’s ability to walk for recreation or transport.²¹ The NEWS-AU comprises 7 subscales: “land-use mix diversity” (ie, how close the participant lives to nonresidential destinations such as shopping centers), “land-use mix access” (ie, how easy it is for the participant to access nonresidential destinations such as shopping centers), “street connectivity” (ie, are there many dead-end streets in the local area?), “infrastructure for walking” (ie, are there many footpaths for walking in the local area?), “aesthetics of area,” “traffic safety,” and “safety from crime.” The mean item scores for the NEWS-AU subscales were calculated.²² The residential density subscale was scored by applying published weightings relative to the average density of a single-storey home to item scores and summating scores within the subscale.²¹

Sample Size

Using t tests for a moderate effect size (d = 0.6) and an α of .05, 35 participants per group would be required to detect moderate difference in steps/day between groups. Allowing for a 30% noncompliance rate, the group sample with NSCLC was increased to 50. We used a moderate effect size to calculate sample size because there were no data available to calculate difference in steps or GPS variables at the time of designing the study.

Statistical Methods

Data were analyzed using SPSS Windows Version 20.0 (SPSS, Chicago, IL). Descriptive statistics and graphical displays were used to identify missing and out-of-range values and to assess the distributional characteristics of test scores prior to formal analysis. Descriptive statistics were also used to assess compliance with assessments and to summarize characteristics and outcome data by group. Pearson’s χ² for nominal variables, Mann–Whitney U tests for ordinal variables, and independent samples t tests for continuous variables were used to compare demographic and clinical characteristics of consenters and study decliners. One-way analysis of covariance was used to assess group differences in continuous outcomes after adjusting for preexisting differences associated with age, provided equal variances could be assumed. If the Levene Statistic was significant, however, group differences were assessed with the Mann–Whitney U test. Alpha was set at .05 (2-tailed) for these analyses. Standard effect size indices (Cohen’s d and r as appropriate) were used to quantify the size of between group differences.²³ Pearson’s correlation coefficient was used to assess bivariate relations between steps/day and both exercise motivation and environmental barriers to PA,²⁴ with coefficient associations interpreted as weak ≤0.25, fair 0.25 to 0.50, moderate 0.50 to 0.75, and large 0.75 to 1.0.²⁵

Results

Between November 2010 and April 2012, 1018 patients were screened for inclusion into the group with NSCLC, of whom 20% (n = 206) were eligible and 9% (n = 90) were approached (Figure 1). The consent rate was 62% (n = 56). The main reason for ineligibility was “diagnosis other than NSCLC” 46% (n = 441), and for nonconsent the main reason was “too distressed/anxious with diagnosis” 23% (n = 8; Figure 1). There was not a statistically significant difference in the age between patients who consented (mean [SD] = 68.5 ± 9.3 years) and those who declined to consent to participate in the trial (mean [SD] = 70.1 ± 10.4 years; mean difference = −1.6; 95% confidence interval [CI] = −6.3 to 3.0; P = .493). There were also no statistically significant differences in gender (χ² = 1.52, P = .218, phi = −0.14), cancer stage (χ² = 9.45, P = .098, phi = 0.35), or physician-rated ECOG-PS (χ² = 4.87, P = .181, phi = 0.26) between consenters and nonconsenters. Six individuals (11%) were withdrawn from this group and not analyzed due to a diagnosis other than NSCLC being confirmed after surgery (Figure 1). During the same time period, 36 healthy individuals expressed interest in the study, all were eligible, and 97% (n = 35) consented. Fifty patients with NSCLC and 35 healthy individuals were included in final analysis (Figure 1).

Figure 1.

Consort diagram.

Accelerometery data were analyzed for 56% (n = 28) of the NSCLC group and 86% (n = 30) of the healthy group. The main reason for the lack of data in the NSCLC group was patient noncompliance (n = 10, 20%) or insufficient time for monitoring to occur before treatment commenced (n = 8, 16%; Table 2). There was a significant difference in the respiratory function measured by forced expiratory volume in the first second (FEV₁) between patients who did have accelerometery data (mean [SD] = 83.2 ± 16.5% predicted) and those that did not have accelerometery data (mean [SD] = 64.7 ± 21.8% predicted; mean difference = −18.5% predicted; 95% CI = −30.7 to −6.3; P = .004). No differences existed for age (P = .260), comorbidities (P = .806), cancer stage (P = .178), self-reported limitations to walking (P = 1.000), or employment status (P = .948) between patients with and without data. All missing accelerometery data from the healthy group were due to malfunctioning of the device (n = 5, 14%) and this occurred at random (Table 2). The percentage of missing items from questionnaires for the NSCLC and healthy groups were 1.11% and 0.07%, respectively, and items were missing at random.

Table 2.

Compliance With Accelerometery Measurement.

	Group 1: Lung Cancer		Group 2: Healthy
Reason	n	%	n	%
Data collected and analyzed	28	56	30	86
No data
Device worn for less than required time	10	20	0	0
Device not worn—participant declined	1	2	0	0
Device not worn—not timed pretreatment	8	16	0	0
Device worn but malfunctioned	0	0	5	14
Inpatient—device not given	1	2	0	0
Other reason	2	4	0	0

As expected, there were significant between-group differences for age (P = .005), living arrangements (P = .042), employment status (P = .001), smoking status/history (P < .001), self-reported limitation to walking (P = .002), FEV₁ (P < .001), and comorbidities (P < .001; Table 3), representing the nature of the “healthy” versus “cancer” cohort.

Table 3.

Demographic Characteristics of Groups.

	Group 1: Lung Cancer (n = 50)		Group 2: Healthy (n = 35)
Variable	n	%	n	%	P Value
Age in years, mean (SD)		68.5 ± 9.3		62.6 ± 8.9	.005*
Gender, male	32	64	19	54	.368
Social situation					.042*
Home alone independent	11	22	3	9
Home with family	33	66	32	91
Home with supports	5	10	0	0
Retirement village	1	2	0	0
Residential location, rural	13	26	3	9	N/A
Employment status					.001*
Working	6	12	19	54
Temporary/permanent sick leave	9	18	0	0
Home duties/not employed/retired	35	70	16	46
Education			N/A
University educated	3	90
Not university educated	28	10
Missing	[19]
Smoking status					<.001*
Never smoker	5	10	18	51
Current smoker	15	30	2	6
Ex-smoker	30	60	15	43
Smoking, median [IQR] pack year Hx	40.0 [19.7-54.2]		0.0 [0.0-2.0]		<.001*
Comorbidity score, mean (SD)	9.3 ± 3.6		4.0 ± 3.7		<.001*
FEV₁, mean (SD) % predicted	74.2 ± 21.2		100.8 ± 12.1		<.001*
ECOG-PS, patient rated					.002*
0	28	56	32	91
1	19	38	3	9
2	3	6	0	0
Self-reported limitation to walking					.002*
Yes	17	34	2	6
No	33	66	33	94
Missing	[1]		[0]
Use of gait aid					.254
No	43	86	35	100
SPS	3	6	0	0
4WF	3	6	0	0
Other	1	2	0	0
Cancer stage
Stage I	21	43
Stage II	10	20
Stage III	18	37
Cancer treatment^a
Radiotherapy palliative <5 fractions	3	6
Radical radiotherapy only	8	16
Chemotherapy and radiotherapy	10	20
Chemotherapy only	1	2
Surgery and chemotherapy	12	24
Surgery only	12	28
Surgery, chemotherapy, and radiotherapy	1	2
No treatment	1	2
Days from assessment to treatment,^b median [IQR] days	12 [4-28]

Abbreviations: 4WF, four wheel frame; Ax, assessment; ECOG, Eastern Cooperate Oncology Group–Performance status; FEV₁, forced expiratory volume in the first second; Hx, history; IQR, interquartile range; n, number; N/A, not assessed; SD, standard deviation; SPS, single point stick.

All participants with lung cancer were tested prior to treatment, treatment listed occurred following outcome measurement.

Time from assessment as part of this study to commencement of chemotherapy/radiotherapy or day of surgery.

Statistically significant.

Healthy individuals were engaged in higher levels of PA than the individuals with NSCLC (Table 4). There was a mean difference (95% CI) between groups of 2363 steps/day (685 to 4040; P = .007; Table 4). Although there was no significant difference in the daily time spent walking outdoors (Table 4; Figure 2), there was a trend for individuals with NSCLC to spend less time outdoors (a mean of 40 minutes less; P = .130; Table 4). There was no difference in the furthest distance travelled away from the home (P = .883). Time spent close to the home, either within 100 m or 1 km from the home, was not significantly different between groups (Table 4). However, the healthy group spent significantly more time outdoors in their local neighborhood area 1 to 5 km from their home (P < .001; Table 4).

Table 4.

Comparison of Results Between Groups^a.

	Group 2: Lung Cancer		Group 2: Healthy
Measure	n	Mean (SE)	n	Mean (SE)	P Value	Mean Difference (95% CI)	Effect Size
Steps per day*	28	6200 (579)	30	8563 (558)	.007*	2363 (685, 4041)	0.79
Time spent outdoors per day (minutes)	24	99.1 (82.7)	29	139.3 (103.7)	.130	40.2 (−12.3, 92.8)	0.43
Depression	44	4.2 (3.3)	35	2.5 (3.0)	.027*	−1.7 (−3.2, −0.2)	0.53
	n	Median (IQR)	n	Median (IQR)	P Value	Effect Size
Outdoor walking per day (minutes)	25	6.7 (0.0-35.2)	29	7.5 (0.0-39.0)	.874	0.02
Furthest distance travelled from home (km)	25	35.1 (6.8-160.9)	29	37.4 (19.9-88.6)	.883	N/A
Time spent <100 m away from home (minutes)	25	4.8 (0.9-7.7)	29	6.2 (3.2-11.3)	.094	N/A
Time spent 100 m to 1 km away from home (minutes)	25	13.6 (8.3-25.6)	29	10.6 (7.6-31.8)	.801	N/A
Time spent 1 km to 5 km away from home (minutes)	25	12.2 (7.0-25.7)	29	43.0 (16.1-63.1)	<.001*	N/A
Time spent >5 km away from home (minutes)	25	36.6 (0.0-105.1)	29	29.3 (6.4-101.4)	.748	N/A
BREQ External regulation	39	0.0 (0.0-0.0)	34	0.0 (0.0-1.0)	.011*	0.30
BREQ Introjected regulation	42	0.0 (0.0-1.0)	35	1.3 (0.7-2.3)	.005*	0.32
BREQ Identified regulation	38	2.0 (1.0-3.0)	35	3.2 (2.5-3.7)	.002*	0.37
BREQ Intrinsic regulation	40	2.0 (1.0-3.0)	35	3.0 (2.2-4.0)	.007*	0.31
BREQ Relative Autonomy Index	37	4.0 (2.5-7.0)	33	6.8 (5.2-7.5)	.001*	0.20
BREQ Amotivation	39	0.0 (0.0-1.0)	34	0.0 (0.0-0.5)	.084	0.40

Abbreviations: BREQ, Behavioral Regulation of Exercise; CI, confidence interval; IQR, interquartile rage; n, number; SE, standard error.

Effect sizes = 0.2 small difference, 0.50 medium difference, 0.80 large difference.²³ Higher regulation scores represent more motivation to exercise due to that regulation. Higher Relative Autonomy Index represents more self-determination to exercise. Higher amotivation scores represent less motivation to exercise.

Statistically significant.

Figure 2.

Daily outdoor walking time (raw data)—comparison between groups.

Individuals with NSCLC performed fewer PA sessions per week than healthy individuals (median [interquartile range; IQR] 4.0 [0.0-7.0] sessions vs 6.0 [2.0-7.0] sessions, P = .024). Thirty-percent (n = 15) of the NSCLC group reported not participating in any PA, compared with only 6% (n = 2) of the healthy group. The most frequent type of PA undertaken during the week was also significantly different between groups (P = .046): the most commonly reported activity in both groups was walking (62%, n = 31, NSCLC vs 69%, n = 24, healthy); however, the group of healthy individuals more commonly reported alternative activities including dancing (0% NSCLC vs 9%, n = 3, healthy), golf (2%, n = 1, NSCLC vs 6%, n = 2, healthy) and cycling (2%, n = 1, NSCLC vs 6%, n = 2, healthy). In addition, the group with NSCLC watched more daily television than healthy individuals (median [IQR] 4.0 [2.0-5.0] hours NSCLC vs 2.0 [1.5-3.0] hours healthy; P = .001).

Individuals with NSCLC had significantly higher depression scores and lower motivation to exercise than healthy individuals (Table 4). No differences existed in amotivation scores between groups, with both groups having low levels of amotivation (ie, no intention to exercise) with median scores of zero (Table 4). In the NSCLC group, higher levels of depression and lower levels of intrinsic regulation (ie, gaining pleasure from exercise) were found in individuals who took fewer steps per day (correlation between depression and steps/day r = .39; correlation between intrinsic regulation and steps/day r = .29). Similarly in the healthy group we found that people who were less active (fewer steps per day) had lower levels of intrinsic regulation (r = .27), and we also found that healthy individuals who were less active had lower levels of identified regulation (ie, valuing the benefits of exercise; r = .29).

There were no differences in the environmental barriers to PA between groups (see the appendix, which includes between-group data for the Neighborhood Environment Walkability Scale domain items). In the NSCLC group, people who had better land-use mix access (ie, easy access to shopping centers) were more active (correlation between land-use mix access and steps/day r = .25). All other correlations between number of steps per day and the domains measured by the Neighborhood Environment Walkability Scale (NEWS-AU) for both groups were weak (r < .25).

Discussion

Our study is novel, being one of the first to use GPS tracking technology for PA measurement in an adult clinical population beyond single or double patient case reports.^26,27 The outcomes we measured, including time spent walking outdoors, time spent outdoors in total, time spent away from the house, and distance travelled from the house, provide important information not only on PA location and context but also on social isolation. Individuals with NSCLC were less physically active prior to commencing treatment than a group of similar-aged healthy individuals. However, we did not find differences in the time spent walking outdoors or the total time spent outdoors between the groups. There was a nonsignificant trend for those with NSCLC to spend 40 minutes less time than healthy individual’s outdoors each day, and a significant reduction in time spent in the local neighborhood area between 1 and 5 km from the home. Although further adequately powered studies are required for confirmation, these findings do provide support for the use of GPS tracking to precisely quantify outdoor activity levels in clinical populations, as this information would be difficult to derive using simple self-reported feedback.

GPS technology provides the ability to measure outdoor activity and outdoor time, which traditional PA measures (such as accelerometers) have been unable to quantify. The ability to measure outdoor time is particularly important given the increasing evidence linking time spent outdoors with improved outcomes, including both physical and mental health.^8,9 Our results demonstrated a significant difference between groups for time spent in the local neighborhood area, and this provides insight into the pattern of outdoor behavior. The healthy group spent ≈3.5 times as long outdoors in this region compared to the group with NSCLC (43 vs 12 minutes), whereas for the other 3 regions there was little difference. This finding may be due to a number of factors, including the group with NSCLC replacing their local neighborhood activities with time spent nearer the location of their clinical appointment (although this was not statistically significant); a loss of desire to travel within the local neighborhood due to the diagnosis (and potentially the stigma associated with the diagnosis of NSCLC); or as a consequence of the loss of mobility associated with the disease, impairing the patients physical ability to leave the home and engage with the community. In addition, the restricted time in the local neighborhood may coincide in patients who are socially isolated. Reduced PA is commonly seen in socially isolated older adults.²⁸ The barriers to outdoor behavior and community engagement that are experienced by patients with NSCLC are potential factors that can be addressed in attempts to enhance health outcomes and activity levels. To our knowledge, this is the first study to use the BREQ-2 to assess exercise regulation in individuals with NSCLC. Our results suggest that amotivation (failure to value exercise) was not different between individuals with NSCLC and healthy individuals; however, all other components of exercise regulation were significantly worse in individuals with NSCLC. Overall, individuals with NSCLC had less self-determination to exercise, which is not surprising since we also demonstrated PA levels were low. The measure of exercise motivation was taken straight after a new diagnosis of NSCLC, and therefore, the distress and knowledge of the new diagnosis may have influenced the measure of an individual’s motivation to exercise. The baseline data reflect motivation to exercise pretreatment and may or may not be different to motivation to exercise 2 weeks prior when the patient did not have the diagnosis. Our results also showed that people who had higher levels of depression were less active. It is not possible from our results to determine causality, and therefore we do not know if less activity increases depression levels or if depression lessens the motivation to be active in this population. In both the healthy older and cancer populations, exercise has the ability to reduce depression severity^9,29,30; consequently, both depression and exercise motivation are important outcomes to target with interventions to increase PA in patients with NSCLC in the future.

Higher levels of PA have a dose-dependent relationship with a decreased risk of developing cardiovascular disease, stroke, hypertension, colon cancer, breast cancer, type 2 diabetes, and osteoporosis in the general population.³¹ Importantly, PA is also a determinant of cardiorespiratory fitness and PA levels are a predictor of premature mortality.^32-34 While no studies have investigated the association between PA and survival in NSCLC, performance status and fitness at time of diagnosis are known, strong and independent predictors of survival in NSCLC.^35,36 Not surprisingly, PA is increasingly being recognized as an important outcome in the NSCLC population as well. Further studies are now needed to investigate if higher levels of PA are associated with improved survival in NSCLC.

There is a growing interest in the role of “prehabilitation” for the cancer population with the aim to optimize health presurgery and therefore accelerate the postsurgical recovery and reduce morbidity and mortality.³⁷ Research findings specifically from the lung cancer population are promising, with studies conducted to date demonstrating improvements in exercise capacity and postoperative respiratory function after a period of preoperative physiotherapy and/or pulmonary rehabilitation.^38,39 However, the studies conducted to date are limited, with small numbers, lack of follow-up, and generally poor methodological (only 2 studies are randomized controlled trials) design, establishing a need for further research in this area. The time of cancer diagnosis is potentially a “teachable moment” when patients may psychologically respond well to exercise interventions. The median [IQR] time from our physical assessment to cancer treatment was 12 days [4-28] (Table 3), and therefore, our work has provided additional evidence that recruitment and measurement of PA outcomes prior to surgery is feasible within the pretreatment “window.” Similarly, PA intervention within this pretreatment window may also be possible for some individuals.

We included a measure of environmental barriers to PA to investigate the relationship they may have with PA in our study. The importance of the environment positively or negatively affecting an individual’s PA behavior is being increasingly recognized, with factors such as better street connectivity, greater residential density, and closer proximity to shopping centers known to encourage increased participation in PA in the general population.⁴⁰ Surprisingly, land-use mix access (intermingling of residential and commercial users) was the only environmental factor to be correlated with steps per day, and this was only found in the group with NSCLC. This result suggests that individuals with NSCLC, who had better access to nonresidential destinations (such as shopping centers), were more active than those who did not. Individuals who live closer to commercial areas (such as shopping centers) are more likely to walk for transport, rather than drive, to the shopping center promoting incidental activity.⁴⁰ Our results suggest this may be the case for people with NSCLC. It is also possible that destinations such as shopping centers where everything is available at one place (shops, food) and the environment is temperature controlled may attract people with impairments since it is easier to walk around inside such spaces.

Our study also demonstrated that individuals with NSCLC watched significantly more television per day than healthy individuals. At time of diagnosis individuals with NSCLC watched a median of 4 hours per day (28 hours/week), which increases their risk of all-cause and cardiovascular disease mortality (≥4 h/week)⁴¹ and is also well above the highest risk category (>14 h/week) for abnormal glucose metabolism and type 2 diabetes.⁴² The increased sedentary behavior of patients with NSCLC, in comparison to healthy individuals, could be due to disease symptoms such as fatigue, physical debility impairing ability to be active, low motivation to exercise, and higher levels of depression. In addition, a greater proportion of the group with NSCLC were not working, increasing the time during the day when they may have watched television (Table 3).

Limitations

This study is limited by the lack of PA data collected. We had problems with malfunctioning of the KinetaMap device in the control group and participant noncompliance. During the piloting stage of the study, performed in 2009, few systems were available for assessing combined triaxial accelerometery and GPS data. The KinetaMap was low-cost and portable enough to use for personal tracking; however, design factors such as it not being water or shock proof may have been a cause for the malfunctions observed. Although this is a limitation with the present study, future similar research could use combined GPS and accelerometer activity monitors, such as Smartphones. Participants who were noncompliant with wearing the KinetaMap device may have been those with lower PA levels based on our evidence that patients who complied had significantly better respiratory function than those who did not. Twenty-six percent (n = 9) of patients who declined to participate in the study did so because they were too distressed with the cancer diagnosis or did not feel up to being involved. Therefore, it is possible that the patients who declined were engaged in even lower levels of PA than those we recruited. However, we found no difference between physical function (measured using the ECOG-PS) of study consenters and those who declined. In addition, caution must be used in concluding that patients with NSCLC are significantly less active than healthy controls given the time period in which this data were collected (individuals with NSCLC had data regarding their PA collected in the short time period between time of diagnosis and treatment commencement).

We did not investigate if the weather had an impact on participation in PA. Previous research has reported that weather affects the amount of walking that individuals with NSCLC engage in, due to fear of falling and a particular dislike of the cold.⁴³ We ensured 1 of the 5 days of monitoring was a weekend day to account for potential differences in PA between week days and weekend days. Reports suggest that number of steps taken on Sundays can drop up to 26% depending on the month of assessment (and weather).⁴⁴ There were slightly more individuals who were currently employed in the healthy group. This could have affected the time spent outdoors (for indoor occupations) or level of PA (for physically active occupations).

Conclusion

Using novel global positioning system tracking and accelerometers we measured physical activity and outdoor behavior in a patient population and healthy cohort. Global positioning system tracking appears to be a feasible methodology for adult cancer patients and holds promise for use in future studies investigating physical activity and/or lifestyle behaviors.

Footnotes

Appendix

Comparison of Environmental Barriers to Physical Activity Between Groups.

	Group1: Lung Cancer			Group 2: Healthy					95% CI
NEWS-AU Domain	n	Mean	SD	n	Mean	SD	P Value	Mean Difference	Lower	Upper	ES^a
Residential density	40	121.5	36.5	35	131.1	23.1	.174	9.6	−4.3	23.4	0.32
Land use mix diversity	43	2.6	0.9	35	2.8	0.7	.155	0.3	−0.1	0.7	0.33
Land use mix access	41	2.9	0.8	35	3.0	0.6	.682	0.1	−0.3	0.4	0.10
Street connectivity	40	2.6	0.7	35	2.8	0.6	.146	0.2	−0.1	0.5	0.33
Walking facilities	43	3.1	0.9	35	3.4	0.5	.610	0.3	−0.0	0.6	0.42
Aesthetics	41	3.2	0.6	35	3.2	0.6	.746	0.0	−0.2	0.3	0.06
Traffic safety	42	2.7	0.5	35	2.8	0.4	.285	0.1	−0.1	0.3	0.25
Crime safety	42	3.1	0.8	35	3.2	0.6	.475	0.1	−0.0	0.4	0.16

Abbreviations: 95% CI, 95% confidence interval; ES, effect size; n, number; NEWS-AU, Neighborhood Environment Walkability Scale–Australian Version; SD, standard deviation.

Effect size = Cohen’s d.

Acknowledgements

The authors would like to thank Selina Parry, Karla Gough, Linda Mileshkin, Meinir Krishnasamy, Andrew Murnane, Maeve Sorohan, Clare Fitzmaurice; the staff of the Austin Hospital, Royal Melbourne Hospital and Peter MacCallum Cancer Centre Physiotherapy and Respiratory Medicine Departments; and the participants for their contribution to the study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by grants from the Victorian Cancer Agency, Australia, and Eirene Lucas Foundation, Australia. Dr. Granger was supported by an Australian Post-graduate Award PhD scholarship and Helen Macpherson Smith Scholarship.

References

Ostroff

Krebs

Coups

. Health-related quality of life among early-stage, non-small cell, lung cancer survivors. Lung Cancer. 2011;71:103-108.

Cooley

. Symptoms in adults with lung cancer. A systematic research review. J Pain Symptom Manage. 2000;19:137-153.

AIHW & Cancer Australia. Lung Cancer in Australia: An Overview (Cancer Series No. 64; Cat. No. CAN 58). Canberra, ACT: Australian Institute of Health and Welfare; 2011.

Dagnelie

Pijls-Johannesma

Lambin

Beijer

De Ruysscher

Kempen

. Impact of fatigue on overall quality of life in lung and breast cancer patients selected for high-dose radiotherapy. Ann Oncol. 2007;18:940-944.

Schmitz

Courneya

Matthews

. ACSM roundtable on exercise guidelines for cancer survivors. Med Sci Sports Exerc. 2010;42:1409-1426.

World Health Organization. Global Recommendations on Physical Activity for Health. Geneva, Switzerland: World Health Organization; 2010:1-58. ISBN: 9789241599979.

Brock

Huang

Fraser

. Low vitamin D status is associated with physical inactivity, obesity and low vitamin D intake in a large US sample of healthy middle-aged men and women. J Steroid Biochem Mol Biol. 2012;121:462-466.

Sugiyama

Leslie

Giles-Corti

Owen

. Associations of neighbourhood greenness with physical and mental health: do walking, social coherence and local social interaction explain the relationships? J Epidemiol Community Health. 2008;62(5):e9.

Barton

Pretty

. What is the best dose of nature and green exercise for improving mental health? A multi-study analysis. Environ Sci Technol. 2010;44:3947-3955.

10.

von Elm

Altman

Egger

Pocock

Gøtzsche

Vandenbroucke

. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007;85:867-872.

11.

Vandenbroucke

von Elm

Altman

. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Epidemiology. 2007;18:805-835. doi:10.1097/EDE.0b013e3181577511.

12.

Oken

Creech

Tormey

. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol. 1982;5:649-655.

13.

Colinet

. A new simplified comorbidity score as a prognostic factor in non-small-cell lung cancer patients: description and comparison with the Charlson’s index. Br J Cancer. 2005;93:1098-1105.

14.

Webber

Porter

. Monitoring mobility in older adults using global positioning system (GPS) watches and accelerometers: a feasibility study. J Aging Phys Act. 2009;17:455-467.

15.

Williams

Weragoda

Paterson

Clark

. Cardiovascular fitness is unrelated to mobility limitations in ambulant people with traumatic brain injury. J Head Trauma Rehabil. 2013;28:E1-E7.

16.

Zigmond

Snaith

. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983;67:361-370.

17.

Markland

. A modification to the behavioural regulation in exercise questionnaire to include an assessment of amotivation. J Sport Exerc Psychol. 2004;26:191-196.

18.

Murcia

de San Román Blanco

Galindo

Villodre

Coll

. Effects of the gender, the age and the practice frequency in the motivation and the enjoyment of the physical exercise. Fitness Perform. 2007;6:140-146.

19.

Peddle

Plotnikoff

Wild

Courneya

. Medical, demographic, and psychosocial correlates of exercise in colorectal cancer survivors: an application of self-determination theory. Support Care Cancer. 2008;16:9-17.

20.

Cerin

Leslie

Owen

Bauman

. An Australian version of the neighborhood environment walkability scale: validity evidence. Meas Phys Educ Exerc Sci. 2008;12:31-51.

21.

Saelens

Sallis

Black

Chen

. Neighborhood-based differences in physical activity: an environment scale evaluation. Am J Public Health. 2003;93:1552-1558.

22.

Leslie

Saelens

Lawrence

. Residents’ perceptions of walkability attributes in objectively different neighbourhoods: a pilot study. Health Place. 2005;11:227-236.

23.

Cohen

. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum; 1988.

24.

Norman

. Likert scales, levels of measurement and the “laws” of statistics. Adv Health Sci Educ. 2010;15:625-632.

25.

Portney

Watkins

. Foundations of Clinical Research Applications to Practice. 2nd ed. Upper Saddle River, NJ: Prentice Hall; 2000.

26.

Barzilay

Noam

Meir

. Assessing the outcomes of spine surgery using global positioning systems. Spine. 2011;36:E263-E267.

27.

Evans

Hanke

Zielke

Keller

Ruroede

. Monitoring community mobility with global positioning system technology after a stroke: a case study. J Neurol Phys Ther. 2012;36:68-78.

28.

Reed

Crespo

Harvey

Andersen

. Social isolation and physical inactivity in older US adults: results from the Third National Health and Nutrition Examination Survey. Eur J Sport Sci. 2011;11:347-353.

29.

Bridle

Spanjers

Patel

Atherton

Lamb

. Effect of exercise on depression severity in older people systematic review and meta-analysis of randomised controlled trials. Br J Psychiatry. 2012;201:180-185.

30.

Schmitz

Holtzman

Courneya

Masse

Duval

Kane

. Controlled physical activity trials in cancer survivors: a systematic review and meta-analysis. Cancer Epidemiol Biomarkers Prev. 2005;14:1588-1595.

31.

Warburton

Charlesworth

Ivey

Nettlefold

Bredin

. A systematic review of the evidence for Canada’s Physical Activity Guidelines for Adults. Int J Behav Nutr Phys Act. 2010;7:39.

32.

Warburton

. Health benefits of physical activity: the evidence. Can Med Assoc J. 2006;174:801-809.

33.

Stofan

DiPietro

Davis

Kohl

Blair

. Physical activity patterns associated with cardiorespiratory fitness and reduced mortality: the Aerobics Center Longitudinal Study. Am J Public Health. 1998;88:1807-1813.

34.

Jones

Eves

Haykowsky

Freedland

Mackey

. Exercise intolerance in cancer and the role of exercise therapy to reverse dysfunction. Lancet Oncol. 2009;10:598-605.

35.

Jones

Eves

Mackey

. Systemic inflammation, cardiorespiratory fitness, and quality of life in patients with advanced non-small cell lung cancer. J Thorac Oncol. 2008;3:194-195.

36.

Blanchon

Grivaux

Asselain

. 4-Year mortality in patients with non-small-cell lung cancer: development and validation of a prognostic index. Lancet Oncol. 2006;7:829-836.

37.

Singh

Newton

Galvao

Spry

Baker

. A systematic review of pre-surgical exercise intervention studies with cancer patients. Surg Oncol. 2013;22:92-104.

38.

Nagarajan

Bennett

Agostini

Naidu

. Is preoperative physiotherapy/pulmonary rehabilitation beneficial in lung resection patients? Interact Cardiovasc Thorac Surg. 2011;13:300-302.

39.

Pehlivan

Turna

Gurses

. The effects of preoperative short-term intense physical therapy in lung cancer patients: a randomized controlled trial. Ann Thorac Cardiovasc Surg. 2011;17:461-468.

40.

Humpel

Owen

Leslie

Marshall

Bauman

Sallis

. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promot. 2004;18:239-242.

41.

Dunstan

Barr

Healy

. Television viewing time and mortality: the Australian Diabetes, Obesity and Lifestyle Study (AusDiab). Circulation. 2010;121:384-391.

42.

Dunstan

Salmon

Owen

. Physical activity and television viewing in relation to risk of undiagnosed abnormal glucose metabolism in adults. Diabetes Care. 2004;27:2603-2609.

43.

Cheville

Dose

Basford

Rhudy

. Insights into the reluctance of patients with late-stage cancer to adopt exercise as a means to reduce their symptoms and improve their function. J Pain Symptom Manage. 2012;44:84-94.

44.

Chan

Ryan

Tudor-Locke

. Relationship between objective measures of physical activity and weather: a longitudinal study. Int J Behav Nutr Phys Act. 2006;3:21. doi:10.1186/1479-5868-3-21.

Physical Activity Measured Using Global Positioning System Tracking in Non–Small Cell Lung Cancer

Abstract

Keywords

Introduction

Methods

Measurement of Secondary Outcomes

Sample Size

Statistical Methods

Results

Discussion

Limitations

Conclusion

Footnotes

Appendix

Acknowledgements

Declaration of Conflicting Interests

Funding

References