Validation of two self-reported physical activity instruments against accelerometer data in patients undergoing lung cancer surgery

ABSTRACT

Introduction

Finding ways of identifying patients with low levels of physical activity after lung cancer surgery would be beneficial when planning and providing interventions aimed at increasing physical activity.

Purpose

To validate two self-reported physical activity instruments against accelerometer data.

Methods

Self-reported physical activity was assessed with the four category One Month Physical Activity Question (OMPAQ) and the International Physical Activity Questionnaire modified for the elderly (IPAQ-E). Objective measurement of physical activity was performed with the Actigraph GT3X+ accelerometer. All measurements were performed three months after lung cancer surgery.

Results

Three months after surgery, 83 patients provided complete physical activity measurements. There were statistically significant correlations between both of the self-reported physical activity assessed by OMPAQ (r = 0.54, p < .01) as well as IPAQ-E (r = 0.50, p < .01) and objectively measured physical activity (steps/day). The correlations were consistently stronger for the higher intensities of physical activity. Both instruments could identify patients not reaching the recommended levels of physical activity.

Conclusion

Both OMPAQ and IPAQ-E give valid information on physical activity after lung cancer surgery, and might be used for screening patients in clinical settings. The OMPAQ provided stronger correlation and specificity than the IPAQ-E, and might be the preferred clinical choice.

KEYWORDS:

• Lung cancer
• physical activity
• surgery

Introduction

Physical activity is defined as any bodily movement produced by skeletal muscles that results in energy expenditure (Caspersen, Powell, and Christenson, 1985). The World Health Organization (WHO) recommends that people living with chronic conditions including cancer survivors should engage in moderate intensity physical activity for a total of at least 150–300 minutes per week, 75–150 minutes of vigorous intensity physical activity per week, or a combination of moderate and vigorous intensity (Bull et al., 2020). The recommendations also state that sedentary time should be limited, and replacing sedentary time with physical activity of any intensity provides health benefits (Bull et al., 2020). It is further recommended muscle-strengthening activities to be performed on at least two days of the week (Bull et al., 2020). Tudor-Locke et al. (2011) proposed that rough equivalents to 150 minutes/week of moderate intensity physical activity are 7100 steps/day for the general population and 4600 steps/day for individuals with disability or chronic illness.

Methods of assessing physical activity can be divided into three categories: 1) Criterion methods such as direct or indirect calorimetry; 2) other objective methods such as pedometers and accelerometers; and 3) subjective methods such as questionnaires, surveys or activity diaries (Steene-Johannessen et al., 2015). Subjective methods are considered easily administered and therefore convenient to use in the clinical setting, although their ability to detect individuals not meeting physical activity recommendations has been described as less accurate compared with accelerometers (Steene-Johannessen et al., 2015).

Lung cancer is both the most frequently diagnosed cancer and the leading cause of cancer deaths among males worldwide (Torre et al., 2015). At diagnosis, individuals with non-small-cell lung cancer engage in significantly less physical activity and are weaker and more depressed than healthy individuals, and their self-reported physical activity declines over time (Granger et al., 2014). After lung cancer surgery, only approximately 25% of patients are sufficiently active according to the American Cancer Society guidelines for cancer survivors, which recommend the same amount of physical activity as the WHO (Lin, Rau, and Lin, 2015; Philip et al., 2013; Rock et al., 2012). In comparison to healthy individuals, patients who have undergone lung cancer surgery are significantly less physically active; they take fewer steps and spend more time being sedentary when measured 6–10 weeks after surgery (Cavalheri et al., 2016). Higher physical activity in lung cancer patients has been positively correlated with physical functioning, role functioning, emotional functioning and global quality of life scores (Bade et al., 2018). Higher physical activity in lung cancer patients has also been correlated with lower dyspnea, depression and pain (Bade et al., 2018).

Finding valid clinically applicable ways of identifying patients with a low level of physical activity would be beneficial when planning and providing interventions aimed at increasing physical activity after surgery (World Health Organization, 2010). One widely used instrument is the International Physical Activity Questionnaire, which is available in a validated version modified for the elderly (IPAQ-E) (Hurtig-Wennlöf, Hagstromer, and Olsson, 2010). The IPAQ-E consists of four questions, which are responded to by reporting the time spent on different intensities of physical activity. The answers are then transformed into categories. A less time- and resource consuming instrument, with potential of being easily implemented in clinical practice, if proven to be valid, is a single item instrument previously used in patients undergoing heart surgery (Jonsson, Urell, Emtner, and Westerdahl, 2014), hereafter called the One Month Physical Activity Question, or OMPAQ, where the patients report their level of physical activity during the previous month by choosing one of four categories. These two instruments were chosen due to their previous use in research and being available in Swedish. The overall aim of this study was to validate two different self-reported physical activity instruments; OMPAQ and IPAQ-E, against objectively measured accelerometer data. A secondary aim was to evaluate the ability of the OMPAQ and the IPAQ-E to identify patients who do or do not reach the WHO recommendations of physical activity.

Methods

Study design

Criterion validation study.

Patients

The sample consisted of patients participating in a previous randomized controlled trial (Jonsson et al., 2019). These patients underwent elective thoracic surgery due to suspected or confirmed lung cancer at the Department of Cardiothoracic and Vascular Surgery, Örebro University Hospital, Sweden, between December 2013 and January 2017. To be included, patients needed to be able to perform a preoperative spirometric lung function test and a walking test, and be able to read and understand Swedish. The primary randomized controlled trial was registered at ClinicalTrials.gov (NCT01961700), and approved by the Regional Ethical Review Board in Uppsala, Sweden (ref: 2013/199). Informed written consent was obtained from each patient.

Surgery and postoperative management

The surgery was performed either by video-assisted thoracoscopy or by open antero-lateral muscle-sparing thoracotomy, according to the surgeon’s preference. At the end of surgery, a single chest tube was placed in the pleural space for drainage. The chest drainage was removed when there was no air leak or excessive volume of pleural effusion. Pain management followed ordinary routines at the clinic, primarily delivered by continuous epidural infusion. Detailed descriptions of mobilization routines and physiotherapy treatment are given elsewhere (Jonsson et al., 2019). The patients were ready to be discharged when all catheters had been removed and the pain was managed by oral administration. The mean length of stay was 5 ± 4 days.

Outcomes

Self-reported and objective measurements of physical activity were conducted three months after surgery. The OMPAQ was used for assessment of self-reported leisure-time physical activity during the previous month. The patients reported their level of physical activity by choosing one of four categories: 1) sedentary; 2) somewhat active; 3) moderate regular exercise (“Moderate active”); and 4) regular activity and exercise (“Exercise”). This was modified from a question in the Swedish National Institute of Health’s national survey (Boström and Nyqvist, 2008), which asks about physical activity during the previous year; in the present study, it was simply modified to assess the previous month. A professional translation of the OMPAQ from Swedish into English is presented in Table 1. The OMPAQ in this form has not been validated previously.

Table 1. Description of the One Month Physical Activity Question (OMPAQ).

The categorical physical activity question (OMPAQ): “How much have you moved about and exerted yourself in your spare time over the last month?”
Sedentary	You spend your spare time mostly reading, watching TV, at the cinema, or doing other sedentary activities. You walk, cycle, or move about in other ways for less than 2 hours a week.
Somewhat active	You walk, cycle, or move about in other ways for at least 2 hours a week, usually without sweating. This also includes walking or cycling to and from work, other walks, heavier household chores, normal gardening work, fishing, table tennis, and bowling.
Moderate active	You exercise regularly 1–2 times per week for at least 30 minutes each time: running, playing tennis or badminton, or doing another activity that causes you to sweat.
Exercise	You spend time running, swimming, playing tennis or badminton, or doing workouts such as keep-fit an average of at least 3 times a week. Each session lasts at least 30 minutes.

The IPAQ-E was used to assess self-reported sedentary time and physical activity during the previous week (Hurtig-Wennlöf, Hagstromer, and Olsson, 2010). The patients reported time spent on different levels of physical activity: sedentary, walking, moderate intensity physical activity, and vigorous intensity physical activity. This self-reported data was transformed into three levels of physical activity (i.e. Low, Moderate, and High) according to a protocol (IPAQ Group, 2005). The IPAQ-E has been validated in older individuals (Hurtig-Wennlöf, Hagstromer, and Olsson, 2010).

Objective measurement of physical activity was performed with an ActiGraph GT3X+ accelerometer (Manufacturing Technology Inc., Pensacola, FL, USA) at a sample frequency of 30 Hz. The output variable was counts per 10 seconds, processed by filtering and summarizing and presented as counts per minute. The accelerometer also provided number of steps. Patients were instructed to wear the device at the waist for one week, at three months after surgery. Wear time validation was conducted according to Choi, Liu, Matthews, and Buchowski (2011). Settings used were: minimum wear time per day 600 minutes; and a minimum of two wear days. Cutoff points according to Barnett, van den Hoek, Barnett, and Cerin (2016) were used to calculate: sedentary time (i.e. 0–100 vertical axis counts/min); time spent on light activity (i.e. 101–1012 vertical axis counts/min); and time spent on moderate and vigorous intensity physical activity (i.e. MVPA, ≥ 1013 vertical axis counts/min). This cut-point has been suggested for older adults and would be appropriate for the studied population. Accelerometer data were processed using the designated software (ActiLife, version 6.11.9). The patients were instructed to complete the self-reported instruments at the end of the week of wearing the accelerometer, thus covering the same time period.

Statistical methods

Continuous data were summarized as mean ± SD or median (1^st quartile-3^rd quartile), and categorical data were calculated as percentages (%). Spearman’s rank correlation coefficient (ρ) was used to assess correlations between self-reported and objectively measured physical activity. In this study, a Spearman’s rho of 0.30–0.50 is considered low correlation, 0.50–0.70 moderate correlation, 0.70–0.90 strong correlation and 0.90–1.0 very strong correlation (Mukaka, 2012). Divergences between observed and expected values of physical activity categories were tested with the χ² test. The sensitivity and specificity of OMPAQ and IPAQ-E was calculated as the categories that were correctly/falsely classified (%). For all analyses, a p-value ≤ 0.05 was considered statistically significant. All statistical calculations were performed in version 25 of IBM SPSS Statistics (IBM Corporation, Armonk, New York, USA).

Results

Of the 132 eligible patients approached for participation in the study, 107 (81%) accepted and were included in the original study (Jonsson et al., 2019). At three months after surgery, 83 (78% of 107) patients provided accelerometer data and complete self-reported physical activity data and were included in the present analysis. Characteristics of the patients are presented in Table 2.

Table 2. Patient characteristics.

Variable	Three months after surgery (n = 83)
Age (years)	68 ± 8
Gender (male/female, % male)	36/47, 43%
BMI (kg/m^2)	26 ± 4
Smoking (never/former/current) (n)	19/56/8
Thoracotomy/VATS	63/20
Length of surgery (min)	127 ± 52
Length of stay (days)	5 ± 4

Data presented as mean ± SD or number (n) of patients. BMI = body mass index. VATS = video assisted thoracoscopic surgery.

At three months after surgery, the majority (58%) of the patients described themselves as Somewhat active according to the four-category OMPAQ (Table 3). Accelerometer measurement showed that the majority of the time, approximately 800 minutes per day, was spent being sedentary. The distributions of subjectively assessed and objectively measured physical activity are presented in Table 3.

Table 3. Distribution of the physical activity intensity three months after lung cancer surgery.

Physical activity measurement	Physical activity category/intensity	Three months after surgery (n = 83)
OMPAQ	Sedentary, n (%)	20 (24%)
	Somewhat active, (n (%)	48 (58%)
	Moderate active, n (%)	8 (10%)
	Exercise, n (%)	7 (8%)
IPAQ-E	Low, n (%)	23 (26%)
	Moderate, n (%)	35 (43%)
	High, n (%)	25 (31%)
Accelerometer	Sedentary time, min/day	797 ± 201
	Time spent on light activity, min/day	125 ± 40
	Time spent on MVPA, min/day	53 ± 34
	Steps/day	4834 ± 2729
	Counts/day	168129 ± 108817

Number (n) of patients at different levels of self-reported physical activity, measured with the OMPAQ and the IPAQ-E, and accelerometer-derived time spent at different intensities of physical activity.

OMPAQ versus accelerometry

Three months after surgery, there were statistically significant correlations between the OMPAQ and objectively measured sedentary time, time spent on light physical activity, time spent on MVPA, steps per day, and counts per day (Table 4). The correlations were low for sedentary time and time spent on light activity, and moderate for time spent on MVPA, steps per day, and counts per day.

Table 4. Correlations between accelerometer measured and self-reported physical activity three months after lung cancer surgery.

	OMPAQ	IPAQ-E
Sedentary per day	−0.23*	−0.17
Light per day	0.41**	0.22*
MVPA per day	0.52**	0.45**
Steps per day	0.54**	0.50**
Counts per day	0.53**	0.42**

OMPAQ = One Month Physical Activity Question. IPAQ-E = International Physical Activity Questionnaire modified for the elderly. MVPA = moderate and vigorous physical activity. Correlations calculated and presented with Spearmans ρ. A ρ of 0.3–0.5 is considered = low, 0.5–0.7 = moderate, and 0.7–0.9 strong. * p < 0.05 ** p < 0.01

The sensitivity of the Sedentary category of the OMPAQ for predicting a total of less than 4600 steps per day was 75%. The specificity of the Moderate active and Exercise categories for predicting a total of more than 4600 steps per day was 93%.

IPAQ-E versus accelerometry

There was a statistically significant correlation between the IPAQ-E and accelerometer data for time spent on light physical activity, time spent on MVPA, steps per day, and counts per day (Table 4). The correlation was moderate for steps per day, and low for light activity, MVPA, and counts per day (Table 4).

The sensitivity of the Low category for predicting a total of less than 4600 steps per day was 76%, and the specificity of the Moderate and High categories for predicting more than 4600 steps per day was 62%.

OMPAQ versus WHO recommendations

Three months after surgery, more than the statistically expected number of the patients reporting themselves as Sedentary were under the threshold of 30 minutes in MVPA and 4600 steps per day, while more than the statistically expected number of patients in the Exercise category reached the threshold of 30 minutes of MVPA and 4600 steps per day (Tables 5 and Tables 6).

Table 5. Observed and expected number of patients achieving more or less than 30 minutes of moderate and vigorous intensity physical activity (MVPA) per day.

Time after surgery	Physical activity instrument	Physical activity category	Less than 30 min MVPA observed/expected	More than 30 min MVPA observed/expected	p-value
3 months	OMPAQ	Sedentary	12/6	8/14	0.001
		Somewhat active	12/14	36/34
		Moderate	0/2	8/6
		Exercise	0/2	7/5
	IPAQ-E	Low	11/6	10/15	0.004
		Moderate	11/10	24/25
		High	3/8	21/16

Observed and expected number of patients in the different categories that reached more or less than 30 minutes of moderate or vigorous physical activity (MVPA) per day. OMPAQ = One Month Physical Activity Question. IPAQ-E = International Physical Activity Questionnaire modified for the elderly. P-value refers to χ² tests.

Table 6. Observed and expected number of patients achieving more or less than 4600 steps per day.

Time after surgery	Physical activity instrument	Physical activity category	Less than 4600 steps/day observed/expected	More than 4600 steps/day observer/expected	p-value
3 months	OMPAQ	Sedentary	15/10	5/10	0.001
		Somewhat active	25/24	23/24
		Moderate	0/4	8/4
		Exercise	1/4	6/4
	IPAQ-E	Low	16/10	5/11	0.005
		Moderate	16/17	19/18
		High	7/12	18/13

Observed and expected number of patients in the different categories that reached more or less than 4600 steps/day. OMPAQ = One Month Physical Activity Question. IPAQ-E = International Physical Activity Questionnaire modified for the elderly. P-value refers to Chi2 tests.

IPAQ-E versus WHO recommendations

Three months after surgery, more than the statistically expected number of patients in the Low category failed to reach 4600 steps per day, while more than the statistically expected number of patients in the High category achieved at least 4600 steps per day (p = .005) (Tables 5 and Tables 6). There were similar results for the threshold of 30 minutes of MVPA per day; more than expected in the Low category failed to reach 30 minutes of MVPA per day and more than expected in the High category reached 30 minutes of MVPA per day (p = .004) (Tables 5 and Tables 6).

Discussion

In this study, the validity of two instruments used for assessment of physical activity, OMPAQ and the IPAQ-E, was tested against objectively measured accelerometer data in patients who had undergone lung cancer surgery. Both instruments correlated statistically with the accelerometer data three months after surgery. The correlation coefficients were consistently higher for the higher levels of physical activity, and both instruments seemed to provide less valid assessment of sedentary behavior, suggesting that both instruments are better at describing physical activity at higher physical activity levels than at lower intensities.

For OMPAQ the strength of correlation was moderate, with increasing correlation for higher intensity of physical activity, ranging from low for sedentary and light activity to moderate for MVPA, steps per day and counts per day were found between the OMPAQ and accelerometer data three months after surgery.

IPAQ-E also showed a correlation against accelerometer data, but the correlation was lower regarding MVPA, steps, and counts. Of the two instruments, the OMPAQ would be the preferred choice in the clinical setting due to better validity as compared to IPAQ-E. The single-item instrument is also easy to handle and does not require any calculations of data collected, as necessary for the IPAQ-E.

Conflicting results have been described regarding validity of the IPAQ in different populations; supporting our results, a study among university students showed that the IPAQ provides valid estimates of physical activity (Rodriguez-Munoz, Corella, Abarca-Sos, and Zaragoza, 2017). Further, Hurtig-Wennlöf, Hagstromer, and Olsson (2010) found that the IPAQ-E provided acceptable estimates of physical activity in elderly individuals, with a specificity of 85% for identifying less-active individuals, and a sensitivity of 81% for identifying more-active individuals. On the other hand, Minetto et al. (2018) found inadequate validity for the IPAQ when compared with pedometry in obese and diabetic patients, and a systematic review concluded that the IPAQ overestimated physical activity, with a mean over-estimate of 106%, when compared to an objective device (Lee, Macfarlane, Lam, and Stewart, 2011). As in our study, the systematic review showed that the IPAQ had stronger validity for the higher levels of physical activity. Our results showed a specificity of 62% and a sensitivity of 76% for IPAQ-E, while the OMPAQ provided stronger specificity (93%) and equal sensitivity (75%), further favoring the OMPAQ as the best clinical choice of the two instruments.

Few of the patients in the OMPAQ Sedentary category achieved 30 minutes of MVPA per day, while all patients in the Moderate active and Exercise categories reached this cutoff point. When this cutoff is used as the definition of being sufficiently physically active, the OMPAQ can identify patients who do not achieve the recommended level of physical activity, and hence could be seen as a clinically applicable instrument.

The threshold of 4600 steps per day as a minimum for meeting the WHO recommended level of physical activity, as proposed by Tudor-Locke et al. (2011) for patients with disabilities, would probably be appropriate for the population in this study; that is, older patients who have undergone surgery. Most of the patients in the Moderate active and Exercise categories, and few of the patients in the Sedentary category, reached 4600 steps per day. When using 4600 steps per day as the cutoff for meeting physical activity recommendations, the OMPAQ Sedentary category identifies patients who are not sufficiently physically active, and the Moderate active and Exercise categories identify patients who are sufficiently physically active. For patients in the Somewhat active category, a more thorough evaluation would be needed to assess whether they are adequately physically active. Further strengthening this, the OMPAQ Sedentary category had a sensitivity of 75%, meaning that 75% of the patients in the category took less than 4600 steps per day. Moreover, the specificity of the Moderate active and Exercise categories was 93%, meaning that 93% of these patients took more than 4600 steps per day.

Few patients in the IPAQ-E Low category achieved 4600 steps per day, while most of the patients in the High category reached this cutoff. The patients in the Moderate category were evenly distributed between reaching/not reaching 4600 steps per day. The sensitivity for identifying patients with less than 4600 steps per day was 76% and the specificity for identifying patients with more than 4600 steps per day was 62%. Thus, the IPAQ-E seems able to identify patients with both insufficient (Low) and sufficient (High) level of physical activity. For patients in the Moderate category, a more thorough evaluation would be necessary. The IPAQ-E seems able to identify the least active patients, and may be used in clinical settings.

There is a time span difference in the two instruments assessed in this study; the OMPAQ covers physical activity during the previous month, while the IPAQ-E covers the previous week. Both instruments seek to describe habitual level of physical activity, as does the one week of accelerometer measurement. The similar results of the measurements and the correlation with accelerometer measurement could be interpreted as meaning that the assessments give valid information on habitual level of physical activity, independent of time span. Which instrument to choose will depend on the population and the setting. The OMPAQ, being a single-item instrument and not requiring any calculations, is quicker to use and provides higher correlation and specificity than IPAQ-E, and would be the preferred choice in the clinical setting. The IPAQ-E has been previously used in research, and so might be the preferred choice in scientific studies in order to enable comparisons with previous findings.

The amount of MVPA may seem high, but this is due to the cutoff value used (Barnett, van den Hoek, Barnett, and Cerin, 2016). This cutoff, which is suggested for older adults, does seem relevant for the studied population, being a sample of older patients who have undergone lung cancer surgery. The sensitivity and specificity were calculated on steps, and the cutoff for MVPA did not have any impact on these results.

Considering the recent recommendations (Bull et al., 2020), which are clearly stressing the importance of avoiding sedentary time, it becomes even more important to identify individuals with low level of physical activity. Physiotherapists have an important mission to support and facilitate physical activity in inactive individuals. The current evaluation of two simple instruments could contribute to this in the clinical setting, where it is important to identify these individuals in order to provide interventions aiming at increasing physical activity to improve health. Our results provide two validated instruments for screening patients who have undergone lung cancer surgery. The instruments might have potential in other patient groups, but this needs to be further validated in future studies.

In conclusion, both instruments might be used with the purpose of finding individuals who do not reach the recommended level of physical activity after lung cancer surgery. The OMPAQ gave stronger correlation and specificity than the IPAQ-E, and might be the preferred choice in the clinical setting.

Acknowledgments

This study was sponsored by grants from the Research Committee of Örebro County Council, the Swedish Heart and Lung Patients National Association, and the Swedish Cancer Society.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the The Swedish Heart and Lung Association; The Örebro County Council; The Swedish Cancer Foundation.