Article Text

Download PDFPDF

How do I measure physical activity in my patients? Questionnaires and objective methods
  1. B E Ainsworth
  1. Department of Exercise and Wellness, Arizona State University, Mesa, Arizona, USA
  1. Dr B E Ainsworth, Department of Exercise and Wellness, Arizona State University, 7350 E Unity, Mesa, AZ 85242, USA; Barbara.Ainsworth{at}


How do healthcare providers know if their patients are getting enough physical activity to promote good health and to reduce their risks of chronic diseases and injury? The first step is to identify the patient’s current level of physical activity using questionnaires and/or motion sensors. Questionnaires assess activity levels by having patients answer a set of questions about the types and amounts of activity performed at some time in the past. Motion sensors assess physical activity by patients wearing a small monitoring device that records their body movement as it occurs. If a provider is interested in determining a patient’s caloric energy expenditure, he/she can apply statistical regression models to the questionnaire and motion sensor data to estimate kilocalories. If more precise measures of energy expenditure are desired, a provider can use the isotopic doubly labelled water method to estimate kilocalories; however, this method is costly and is impractical in non-research clinical settings.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

To understand the breadth of physical activity and energy expenditure assessment methods, it is useful to view a schematic of the construct of “movement”.1 As shown in fig 1, physical activity is a behaviour that results in caloric energy expenditure. Although many use these terms interchangeably, they are distinctly different and are measured differently. Direct methods measure movement as it occurs, such as a pedometer counting steps taken or the doubly labelled water method that measures hydrogen and carbon utilisation to assess energy expenditure. Indirect methods provide indicators of physical activity and energy expenditure, such as the use of questionnaires for patients to recall characteristics of recent activity or the measurement of heart rate, a physiological correlate of energy expenditure. Because many providers are interested in encouraging their patients to increase physical activity, this discussion will focus on practical direct and indirect measures that can be used in clinical settings. More detailed reviews of physical activity assessment methods are available elsewhere.24

Figure 1 Conceptual framework for defining and assessing physical activity (PA) and energy expenditure.1


Direct methods are preferred for optimal accuracy because they do not involve elements of recall in the assessment of physical activity; however, they require the use of instruments and/or complex measurement systems that may reduce their practicality in clinical settings.

Motion sensors

Two types of motion sensors are used to monitor physical activity: accelerometers and pedometers. Accelerometers are battery operated electronic motion sensors that reflect the speed and intensity of the body’s movement. Accelerometers are typically worn at the waist, with most brands measuring acceleration in the vertical plane Activity data are expressed as counts, with increasing counts reflecting increasing amounts and intensities of movement. Whereas a downside to accelerometers is their expense (approximately US$350) and the need for a technician to download and interpret the data, they have become increasingly popular to assess physical activity. The ActiGraph (Fort Walton Beach, Florida, USA) is used most frequently in research and in surveillance settings to determine the time spent in sedentary, light, moderate and vigorous-intensity behaviours. Used in the 2003–4 US National Health and Nutrition Examination Survey to identify the activity patterns of 6329 children, adolescents and adults,5 the ActiGraph showed that over half of children aged 6–11 years were active for 60 minutes or more per day in moderate and vigorous-intensity activity; however, fewer than 10% of adolescents aged 12–18 years were active at this level. Among adults, fewer than 5% met public health guidelines6 7 to accumulate at least 30 minutes of moderate-intensity activity per day in continuous bouts of 8–10 minutes or longer. In addition, nearly half of adults spent at least 8 h per day in sedentary behaviours, with teens and adults aged 60 years and older being the least active.8

Pedometers have gained widespread popularity in research and practice settings to monitor steps taken in free-living settings. Pedometers are small devices worn at the waist and are triggered during walking. Whereas most people use the pedometer to count their steps without regard for the distance covered, an estimate of distance walked can be obtained by calibrating the pedometer to one’s stride length. Pedometers were first used by Japanese walking clubs and were referred to as “10 000 steps meters” (in Japanese, manpo-kei).9 In 1993, Hatano10 equated walking 10 000 steps per day with 300 kilocalories energy expenditure, a daily amount identified by Dr Ralph Paffenbarger in the College Alumnus Health Study as optimal to reduce the risks of a first heart attack.11 Based on an equivalence of approximately 2000–2500 steps per mile (or 100 steps per minute),12 brisk walking for 30 minutes per day translates to approximately 3000–4000 steps. Using the available literature, Tudor-Locke and Bassett13 identified the step counts as shown in table 1 to characterise activity levels.

Table 1 Table 1 Tudor-Locke and Bassett step count

Pedometers are effective as motivational devices to increase physical activity levels. In a review of eight randomised controlled trials and 18 observational studies lasting an average of 18 weeks, Bravata et al14 noted that pedometers were effective in increasing physical activity levels 2200–2500 steps per day over baseline. Additional steps were related to favourable changes in body mass index and blood pressure. Individuals with a goal of taking 10 000 steps per day had the greatest increase in steps compared with those with other step goals. There is considerable variability in the brands of pedometers available, with some having higher step accuracy than others. For healthy, ambulatory adults, the New-Lifestyles NL-2000, Yamax Digi-Walker SW-200 and the Yamax Digi-Walker SW-701 pedometers have the best accuracy for steps taken.15 For older adults living in assisted-living facilities, the StepWatch 3 Step Activity Monitor is most accurate for counting steps.16 An advantage of using pedometers to promote physical activity in a clinical setting is the relatively low cost of the instrument (approximately US$20) and the ease of use. A disadvantage is that they can be lost, do not store data to record daily activity levels for later retrieval and they may be less accurate when used by people with altered gait patterns.


Indirect methods are used most frequently to assess physical activity in clinical settings to identify a patient’s past activity patterns and to assess detailed information about the types of activity performed.

Records and log books

Physical activity records and logs provide a detailed account of activity types and patterns that are written into a record or diary format. Records and log books can identify the type (eg, gardening, walking, lifting boxes), purpose (eg, home maintenance, transportation), duration (eg, minutes), self-rated intensity (light, moderate, vigorous), and body position (reclining, sitting, standing, walking) for activities completed within a defined observation period (typically several days to a week). Seasonal records (eg, winter, summer) can be kept to obtain information about habitual activity levels and patterns and related seasonal variations in these behaviours. The detail recorded ranges from writing down each activity and its associated duration as it is completed17 to recording activities performed at specified time intervals (eg, every 15 minutes).18 To identify the intensity of the activity, the compendium of physical activities may be used to assign a metabolic energy cost (MET level) to each activity, with a MET equal to the ratio of the activity metabolic rate to the resting metabolic rate.19 Energy expenditure for each activity can be estimated using the following equation:

Kilocalories  =  MET × hours of activity × body weight in kilograms

Although records and logs provide a detailed and comprehensive method of assessing activities, they have a high administrative burden for patients and clinicians. Records and logs may thus best be suited for use with individuals considered to be at high risk of energy imbalance (ie, obesity) or when detailed information is needed about patterns of physical activity for health conditions.


Self-report questionnaires are frequently used to classify physical activity levels and may be sent by mail, self-administered in a clinic setting, or delivered by telephone. Most questionnaires are self-administered; however, with some population groups, such as the elderly, those with low literacy, or children with limited ability to recall details about past physical activity, interview methods may be preferred. Three types of questionnaires have been classified (table 2).

Table 2 Examples of global, recall and quantitative history physical activity questionnaire items

Global questionnaires are brief surveys about one’s general physical activity level. The questionnaires are easy to complete and score, taking less than a minute, and are best suited to obtain a brief classification of activity levels, such as a rating of inactive or active.20 21 Global questionnaires do not provide details about specific types and patterns of physical activity and should not be used if clinicians are interested in identifying whether patients are meeting physical activity guidelines. Recall questionnaires are longer than their global counterparts (7–20 items) and identify details about the frequency, duration and types of activity performed during the past day, week, or month. Summary scores include ordinal scales (eg, 1–5 representing low to high levels of physical activity)22 or continuous data expressed as kilocalories per day or MET-minutes per day.23 Recall questionnaires can reflect various domains of activity, such as exercise, leisure activity, lawn and garden, occupation, transportation, and family care activities, and they can be used to determine whether a patient meets public health activity guidelines. Although used frequently, recall questionnaires have several limitations associated with measurement error: they may be subject to recall bias with time spent in vigorous-intensity activities overestimated;24 habitual, daily activity, such as walking, is difficult to recall and may be underestimated25 and responses may have biases based on demographic (eg, age, sex, education) characteristics of the respondents.26 Quantitative histories are long (eg, approximately 50 items), detailed surveys about the frequency and duration of activities performed over the past year27 or lifetime.28 Activity scores are usually expressed as a continuous variable (eg, kcal/kg per minute), allowing for the evaluation of activity-related energy expenditure. Quantitative activity histories are useful in settings in which clinicians are interested in identifying activity patterns over long periods of time to assess their relation with chronic disease health outcomes. Because they take a considerable time to administer, quantitative history questionnaires are unlikely to be used in clinical settings. This description is thus included for completeness of the discussion about physical activity questionnaires.

Measures of physical inactivity

There is considerable interest in identifying the time spent in sedentary behaviours.29 Sedentary behaviours include working at desk jobs, travelling in cars, watching television, using a computer, reading or doing other inactive pursuits. While some inactivity time is less modifiable than others, such as work or driving long distances, patients can be encouraged to sit less and walk more when possible (eg, walking to meetings, walking during breaks). The treadmill desk has been suggested as a way to replace sitting time with continuous, slow walking during work ( Levine and Miller30 noted that walking at a self-selected speed of 1.1 mph at a vertical (treadmill) workstation expends approximately 100 kcal per hour more than doing similar work while sitting.30 Other inactive behaviours, such as watching television, are more easily replaced with active pursuits. Inactive time can be detected using accelerometers that record sustained minutes with low activity counts, pedometers that show fewer than 5000 daily steps, and with records, logs and questionnaires that identify the frequency and duration of inactive behaviours.2 31 There are patients who may engage in moderate or vigorous-intensity physical activity for 30 minutes or longer per day who also have prolonged bouts of inactivity. This group is a minority, as most people fail to get sufficient levels of activity per day at recommended levels.5 8

In summary, several direct and indirect methods are available to measure physical activity and its related energy expenditure in clinic patients. These measurements exist along a continuum of precision, cost and administrative burden. For most patients, physical activity questionnaires and pedometers will be sufficient to assess their general level of free-living physical activity. These methods are the least expensive and least burdensome, yet still allow for a detailed assessment of physical activity for physical activity promotion efforts. Sophisticated measures of physical activity (accelerometers) and energy expenditure (doubly labelled water) provide the most precision but are not practical for use in most clinical settings.

Clinicians are encouraged to assess their patients’ physical activity routinely, similar to the way they measure body weight, temperature and blood pressure with each clinic visit. This will provide an ongoing record of a patient’s current physical activity level for recommendation of ways to achieve regular, health-enhancing physical activity.



  • Competing interests: None.