`
`ResearchDi est
`
`Series 3, No. 17
`
`June 2002
`
`Taking Steps Toward Increased Physical Activity:
`Using Pedometers to Measure and Motivate
`
`Introduction
`The U.S. Surgeon General ended 2001 with a call to action focused on preventing and decreasing the growing epidemic of
`overweight and obesity that threatens the health and welfare of our nation (U.S. Department of Health and Human Services, 2001).
`The prevalence of obesity (defined for adults as a body mass index ≥ 30 kg/m2) increased from 12.0% in 1991 to 17.9% in 1998
`(Mokdad, Serdula, & Dietz, 1999). In 2000 the prevalence of obesity was 19.8%, providing additional evidence of a continuing
`epidemic trend (Mokdad et al., 2001). Overweight (defined for adults as a body mass index between 25 and 30 kg/m2) and obesity
`are associated with a host of disabilities and chronic illnesses (including diabetes and cardiovascular diseases); approximately
`300,000 annual deaths are attributable to obesity (Allison, Fontaine, Manson, Stevens, & VanItallie, 1999).
`One of the factors that has contributed to the obesity epidemic is physical inactivity (i.e., a sedentary lifestyle). Physical activity is
`one of the ten Leading Health Indicators identified by Healthy People 2010 (U.S. Department of Health and Human Services,
`2000), supporting its importance as a pressing public health issue. Recent national estimates of leisure-time physical activity
`(derived from the 1998 Behavioral Risk Factor Surveillance System; BRFSS) indicate that approximately 50% of American adults
`are not sufficiently active to achieve health benefits; 29% are not physically active at all (Centers for Disease Control and
`Prevention, 2001). Self-reported participation in leisure-time physical activity has remained relatively stable (Centers for Disease
`Control and Prevention, 2001) even amidst increasing trends for overweight and
`obesity of epidemic proportions. Regardless, there has been a noticeable transition in
`work-related physical activity demands (moving increasingly from physical labor to
`sedentary occupations) and also in short-distance transportation modes and patterns.
`For example, U.S. transportation surveys indicate that there has been an annual
`increase in the number of personal vehicles at the rate of approximately 1.5 times the
`population growth and that the average household traveled about 4,000 more miles in
`1995 than in 1990 (U.S. Department of Transportation, 1999). Further, there was a
`37% decline in the number of trips made by children by foot or by bicycle between
`1977 and 1995 (McCann & DeLille, 2000). Together, the estimated direct costs of
`inactivity and obesity account for approximately 9.4% of U.S. health care
`expenditures (Colditz, 1999). Booth (2002) reported that the direct and indirect costs
`of sedentary living is $150 billion. In 1996 the U.S. Surgeon General (U.S.
`Department of Health and Human Services, 1996) endorsed public health
`recommendations (Pate et al., 1995) that individuals minimally strive to accumulate
`30 minutes or more of moderate intensity activity (like brisk walking) on most, if not
`all, days of the week. Since that time researchers and practitioners have likewise been
`striving to implement this recommendation and practically evaluate it. The purpose of
`this article is to explore the potential of a simple pedometer for both measurement and
`motivation.
`
`Published quarterly by the
`President’s Council on
`Physical Fitness and Sports
`Washington, D.C.
`★
`Guest Author:
`Catrine Tudor-Locke, Ph.D.
`Department of Exercise and Wellness,
`Arizona State University East,
`Mesa, Arizona;
`Canadian Centre for
`Activity and Aging,
`The University of Western Ontario,
`London, Ontario, Canada
`★
`
`Co-edited by:
`Drs. Charles B. Corbin and
`Robert P. Pangrazi,
`Arizona State University, and
`Dr. Don Franks,
`University of Maryland
`
`A Brief History of Pedometry
`Although the invention of the pedometer is commonly attributed to U.S. President
`Thomas Jefferson, drawings from the 15th century indicate that Leonardo da Vinci
`was the conceptual originator (Gibbs-Smith, 1978). His early design appeared to be a
`gear-driven device with a pendulum arm designed to move back and forth with the
`swinging of the legs during walking. Thomas Jefferson did enjoy the use of a
`pedometer he purchased in France, however, and likely introduced it to America
`(Wilson & Stanton, 1999).
`
`IPR2017-01058
`Garmin EX1012 Page 1
`
`
`
`Pedometers have been used in Japan to assess physical
`activity and increase walking behaviors for over 30 years
`(Hatano, 1993). Hatano (Hatano & Tudor-Locke, 2001)
`reported that a pedometer came onto the commercial market
`in 1965 under the name of manpo-meter (manpo in Japanese
`means 10,000 steps). Both the slogan and the pedometer
`were widely accepted by the public and organized walking
`clubs seized the concept. Hatano reported that surveys
`conducted at walking events in Japan indicate that >90% of
`respondents have been aware of the slogan for more than five
`years and each household reports ownership of almost 2
`pedometers! At least 10 Japanese-language articles are
`currently listed in PubMed (an electronic search engine).
`Using other search strategies (communicating with Japanese
`collaborators, translating reference sections of held Japanese-
`language articles, translating references obtained through
`Japanese-language search engines) yields another 48
`promising articles. Currently, Japanese-language literature
`about pedometer-assessed physical activity represents an
`untapped source of scientific and practical information.
`Unfortunately, without translation, the contents of these
`articles are inaccessible to most North American researchers
`and practitioners. Support for the translation and review of
`previously inaccessible Japanese-language literature will
`likely contribute much to our understanding and use of the
`pedometer for multiple practical purposes.
`Early English-language research studies used mechanical
`pedometers (Bassett et al., 1996) that were subject to large
`error making them unsuitable as precise research instruments
`(Blair, 1984). Further, researchers were quick to dismiss the
`utility of pedometers based on the poor performance of single
`brands (Tryon, Pinto, & Morrison, 1991). The new generation
`of electronic pedometers is more accurate for recording
`walking-related activity (Bassett et al., 1996). Set against the
`backdrop of a continuing obesity epidemic, and combined
`with the increased emphasis on accumulated moderate
`activity (Pate et al., 1995) as endorsed by the U.S. Surgeon
`General (U.S. Department of Health and Human Services,
`1996), the stage has been set for rapid acceptance of the
`pedometer for both measurement and motivation applications.
`
`Measurement
`How do we measure physical activity in free-living
`populations?
`Assessment of physical activity in free-living conditions (i.e.,
`in the real world) is important to researchers and
`practitioners interested in surveillance, screening, program
`evaluation and intervention. Typically, self-report
`questionnaires, diaries and logs have been used to assess
`physical activity. The limitations to self-report include recall
`bias (Ainsworth, Sternfeld, Slattery, Daguise, & Zahm, 1998;
`Sallis & Saelens, 2000), differential interpretations of terms
`(e.g., light, moderate, vigorous activity) (Wilcox et al.,
`2001), floor effects (the lowest score available is too high for
`some respondents) (Tudor-Locke & Myers, 2001a), and a
`lack of sensitivity to ambulatory activity or walking
`(Ainsworth, Leon, Richardson, Jacobs, & Paffenbarger,
`1993; Kriska et al., 1990; Richardson, Leon, Jacobs,
`Ainsworth, & Serfass, 1994). Ironically, walking is perhaps
`the most important activity to accurately assess; it is
`
`fundamental to all our daily activities and is consistently
`reported as a preferred leisure-time activity choice (Tudor-
`Locke & Myers, 2001a).
`Although self-report approaches to measuring physical
`activity are still considered important to understanding
`context and patterns, there is increasing interest in objective
`monitoring of daily physical activity using electronic motion
`sensors, including accelerometers (Westerterp, 1999a,
`1999b) and pedometers (Rowlands, Eston, & Ingledew,
`1997; Tudor-Locke & Myers, 2001a, 2001b). Both types of
`motion sensors are small, light-weight, unobtrusive
`instruments that are typically worn comfortably at the waist
`and count movement. Accelerometers can detect movement
`in one plane (uniaxial, typically the vertical plane) or up to
`three planes (triaxial). Uniaxial accelerometers typically
`contain a horizontal lever arm with an electronic sensor
`sensitive to distortions in the vertical (up and down) plane.
`The accelerometer records “activity counts” (raw or pure
`movement data) that are the product of frequency and
`intensity (inferred from velocity) of movement sampled at
`set intervals (e.g., over one minute). The results are then
`either displayed as an accumulated total or, more often,
`downloaded for computer analysis. In contrast, the
`pedometer is much simpler in design and requires no
`additional software or expertise to access or interpret data.
`The internal mechanism of a pedometer typically includes a
`horizontal, spring suspended lever arm that moves up and
`down with normal ambulation (e.g., walking, running). An
`electrical circuit closes with each movement detected and an
`accumulated step count is displayed digitally on a feedback
`screen. Pedometers do not, however, record velocity of
`movement, restricting their use to measures of total
`accumulated steps/day, or accumulated steps taken over a
`specific time frame (e.g., during physical education class).
`Some of the newest pedometers count “time in activity.”
`Pedometers with a time feature have a clock that starts with
`the initiation of stepping and stops with inactivity. We must
`await research findings to determine the value of this feature.
`Pedometers display good agreement with accelerometers
`(r=0.80-0.90) (Bassett et al., 2000; Kalscheuer, 2002;
`Leenders, Sherman, & Nagaraja, 2000), indicating that the
`two types of motion sensors measure approximately the same
`total accumulated daily activity. The cost of accelerometers
`($50-400 per unit), costs for additional software and
`calibration hardware, and the associated demands of
`personnel expertise and time, make widespread use of
`accelerometers prohibitive outside the realm of research
`(Tudor-Locke & Myers, 2001a). More and more, researchers
`are beginning to acknowledge, that in terms of practicality,
`pedometers offer the better solution for a low cost ($10-30
`per unit), objective monitoring tool that is accessible to both
`researchers and practitioners (Bassett, 2000; Freedson &
`Miller, 2000; Welk, Corbin, & Dale, 2000; Welk, et al.,
`2000). A common measurement tool and collection protocols
`would help bridge the gap between research and practice.
`
`What do pedometers measure?
`Most pedometers record and display movement as steps
`taken (a simple, raw or pure measure of ambulatory activity).
`Some also have features to estimate energy expended (kcals)
`
`2
`
`IPR2017-01058
`Garmin EX1012 Page 2
`
`
`
`and/or distance traveled (miles or kilometers). Typically the
`user must manually enter a number of variables (including
`any manufacturer-defined combination of gender, stride
`length, weight and/or age) into the pedometer’s program in
`order to obtain a running estimate of caloric expenditure or
`distance traveled. The manufacturer’s actual mathematical
`formula used is usually proprietary, meaning it is not readily
`available for public or scientific scrutiny. Further, this
`process of manipulating the raw step data introduces possible
`error. Individuals with shorter stride lengths appear to do less
`activity (for the same number of steps taken) if only their
`distance traveled is compared to that of individuals with
`longer stride lengths (Bassey, Dallosso, Fentem, Irving, &
`Patrick, 1987; Saris & Binkhorst, 1977). Similarly, reporting
`physical activity as energy expenditure makes it appear that
`obese people are more active than those who are normal
`weight (Tudor-Locke & Myers, 2001b). In agreement with
`this, studies have shown that pedometers are most accurate at
`measuring steps taken (Bassett et al., 1996; Hendelman,
`Miller, Baggett, Debold, & Freedson, 2000), less accurate at
`estimating distance traveled (Bassett et al., 1996; Hendelman
`et al., 2000), and even less accurate at estimating energy
`expenditure (Bassett et al., 2000). For these reasons,
`researchers (Rowlands et al., 1997; Tudor-Locke & Myers,
`2001b) have recommended that steps taken, or steps/day be
`universally adopted as a standard unit of measurement for
`collecting, reporting, and interpreting pedometer data.
`
`Pedometers are not perfect and there are a few potential
`threats to their validity. Although it has been suggested that
`riding in motorized transport may contribute to erroneously
`detected “steps” (Schonhofer, Ardes, Geibel, Kohler, &
`Jones, 1997), the magnitude of the error is approximately 2-
`3% of daily accumulated steps taken and therefore can be
`considered minor (Tudor-Locke, Jones, Myers, Paterson, &
`Ecclestone, in press). Pedometers consistently show more
`error during slow walking (Bassett et al., 1996; Hendelman
`et al., 2000). Specifically, research has shown that the Yamax
`pedometer underestimates (by approximately 25%) steps
`taken at walking speeds of < 60 meters/minute (Bassett et
`al., 1996; Hendelman et al., 2000). Hendelman et al. (2000)
`suggest that this speed of walking is much slower than
`normal self-selected walking speeds and should therefore not
`be considered an important source of error in free-living
`general populations. However, slow, shuffling, gaits
`characteristic of frail and/or institutionalized older adults
`may not be easily detected (Wilcox, Tudor-Locke, &
`Ainsworth, 2002); this is at least one population group for
`whom pedometers may prove to be inappropriate. A
`dissertation from the University of Waterloo provides
`preliminary evidence that this is true (Cyarto, 2001).
`Concern has also been raised about error related to increased
`obesity (Schmalzried et al., 1998; Shepherd, Toloza,
`McClung, & Schmalzried, 1999). In particular, abdominally-
`distributed adiposity may interfere with accurate detection of
`steps taken due to inappropriate placement (e.g., rotation of
`the pedometer horizontally), gait abnormalities associated
`with extreme obesity, and/or a dampening effect. This is an
`important research question that needs to be addressed: is
`there a BMI cut point above which pedometer error is
`unacceptable? Until such information is available, it would
`behoove researchers and practitioners to assess the
`
`pedometer’s validity on each participant during a brief
`walking trial (McClung, Zahiri, Higa, Amstutz, &
`Schmalzried, 2000).
`
`What is the best pedometer to use?
`A number of electronic pedometers are commercially
`available; one has only to type in the keyword “pedometer”
`on any internet search engine to view a variety of
`instruments. Unfortunately, only one study has conducted a
`head-to-head comparison of different brands (Bassett et al.,
`1996). The most accurate brand in that study, the Yamax
`Corporation (Model SW-500, Tokyo, Japan) recorded within
`1% of all steps taken under controlled conditions (walking
`on a 4.88km sidewalk course). Unfortunately, as is often the
`case with consumer items, this particular model has been
`discontinued (Bassett, 2000). It appears that identifying a
`single brand and model for standard use is futile since access
`to specific models is governed by distribution channels and
`we can anticipate continual product development. Therefore,
`before using any particular brand or model of pedometer,
`researchers and practitioners should quickly validate their
`units against the obvious field criterion standard of observed
`steps taken (Tudor-Locke & Myers, 2001b). Simply walk a
`short distance at a normal walking pace wearing the
`pedometer as specified by the manufacturer and
`simultaneously count actual steps taken. Vincent and Sidman
`(in press) conducted a 100-step walk test and also a “shake
`test.” The shake test involved shaking the pedometers in the
`manufacturer’s shipping box 100 times and then recording
`the counts on each pedometer. These researchers reported
`that the percent error for the walk test was <2% and for the
`shake test was <1%. No pedometer exceeded 5% error (i.e., 5
`steps out of 100) on any of the tests. Researchers and
`practitioners should expect similar error when validating
`their own pedometers using similar methods.
`
`How do you collect pedometer data?
`Both researchers and practitioners will be interested in
`collecting pedometer data for screening and evaluation
`purposes. The universal adoption of standardized data
`collection methods and protocols is necessary to bridge the
`gap between research and practice. These methods must not
`only be based on empirical research but must also be feasible
`and acceptable under free-living conditions to researchers,
`practitioners, and their participants. An initial attempt has
`been made to assemble these methods and protocols (Tudor-
`Locke & Myers, 2001b) and they are summarized here with
`some updated information.
`What unit of measure should be used? As previously
`discussed, steps taken (over a defined time period), or
`steps/day is the most appropriate unit of measure. This
`format (rather than energy expenditure or distance traveled)
`will facilitate comparisons between studies and with existing
`programs.
`How long must the pedometer be worn? In the case of
`estimating habitual or customary activity, the monitoring
`frame is the necessarily minimum amount of time that
`participants must wear the pedometer. This has not been well
`established yet and may vary depending upon the
`characteristics of the target population under study. To date,
`
`3
`
`IPR2017-01058
`Garmin EX1012 Page 3
`
`
`
`monitoring frames have ranged from one day (Kashiwazaki,
`Inaoka, Suzuki, & Tamada, 1985) to 14 consecutive days
`(Tryon, Goldberg, & Morrison, 1992). Meager reliability
`evidence has been put forth as yet. Gretebeck and Montoye
`(1992) are often cited to support a monitoring frame of five
`to six days (including weekend days) of pedometer data
`collection with less than 5% error. That study was conducted
`with a sample of young males purposefully recruited on the
`basis of their varied physical activity pursuits. This length of
`monitoring frame may not be necessary in all populations
`especially those considered to be typically sedentary. For
`example, a couple days may be all that is necessary to obtain
`a reliable estimate of habitual activity in individuals living
`with chronic illness or who otherwise participate in few and
`relatively unvaried physical activities (Schonhofer et al.,
`1997; Tudor-Locke, 2001). Until better information is
`available, it is prudent to conduct pilot work in the intended
`population (Tudor-Locke & Myers, 2001b). However, if the
`purpose of the study is to examine cyclical patterns of daily
`physical activity (e.g., associated with day of the week), or to
`promote individual awareness of personal patterns of daily
`physical activity as part of a behavior modification program,
`then a full week (or more in the case of intervention) of
`continuous monitoring may be most effective.
`How should step count results be recorded? When it comes
`down to the specifics of actual data recording, there are two
`obvious choices: either the researcher/practitioner records the
`data (implying the pedometer is sealed and no feedback is
`accessible to the participant) or the participant does
`(implying the pedometer is unsealed and feedback is
`accessible). Regardless of the data recording specifics, there
`is always a concern that participants will alter their behavior
`simply because they are being monitored (also know as
`reactivity). Vincent and Pangrazi (2002) recently ruled out
`reactivity in children wearing sealed pedometers. The
`potential for reactivity using unsealed pedometers has not
`been well explored yet. A thesis at Arizona State University
`focused on this problem found preliminary evidence that
`children did not alter their behavior when monitored by
`unsealed pedometers compared to sealed ones (Ozdoba,
`2002). At this time we do not know if reactivity is a problem
`with adults, regardless of whether or not the pedometer is
`sealed. Additional research is needed to address these
`niggling issues.
`Reactivity aside, a number of recent studies have been
`conducted where the pedometer was unsealed and
`participants took an active role in recording data (Moreau et
`al., 2001; Speck & Looney, 2001; Sugiura, Kajima, Mirbod,
`Iwata, & Matsuoka, 2002; Tudor-Locke, 2001; C. Tudor-
`Locke et al., 2001; Tudor-Locke, Jones, et al., in press;
`Wilde, Sidman, & Corbin, 2001). The practical appeal of this
`process is undeniable; successful self-monitoring opens up
`additional possibilities including surveillance that capitalizes
`on the existing postal system. At least two studies have relied
`on participants returning self-monitored pedometer data in
`this manner (Sequeira, Rickenbach, Wietlisbach, Tullen, &
`Schutz, 1995; Tudor-Locke et al., 2002). On the whole, it
`appears that (given simple instructions), few adults have
`problems recording their total daily steps on a calendar and
`re-setting the pedometer to zero in preparation for a
`subsequent day of data collection. Although it seems likely
`
`(especially with supervision), less information is available on
`children’s ability to take part in their own data collection.
`
`How many steps do people take?
`To date, no single study has yet been conducted to obtain
`representative data on a random population sample.
`Normative data for expected values of steps/day has
`necessarily been assembled from original studies scattered
`throughout the published literature. A systematic review
`(Tudor-Locke & Myers, 2001b) of these studies (32 in total)
`suggests that we can expect between 12,000-16,000
`steps/day in 8-10 year old children (lower for girls than
`boys); between 7,000-13,000 steps/day in healthy younger
`adult samples (lower for women than for men); between
`6,000-8,500 steps/day in healthy older adult samples; and
`between 3,500-5,500 steps/day in individuals with
`disabilities and chronic diseases. Since that time a number of
`additional studies have been conducted. A study of 700+ 6-
`12 year old children reported that girls took between 10,479-
`11,274 steps/day and boys took 12,300-13,989 steps/day
`(Vincent & Pangrazi, in press). Another study of 600+
`adolescents (14-16 year old) also reported values of 11,000-
`12,000 steps/day (again, lower for girls than boys) (Wilde,
`2002). Although the evidence is currently fragmented,
`patterns of pedometer-determined physical activity are
`discernable. Figure 1 presents a summary of these expected
`values. Expected values of steps/day can serve as
`benchmarks for interpreting change and comparison
`purposes but should not be misinterpreted as
`recommendations for appropriate activity levels since we
`may discover that optimal indices associated with important
`health outcomes are higher! Recommendations can only be
`made once the totality of accumulated evidence supports
`specific health-related cut points or indices. This last point
`will be discussed in more detail later.
`
`Motivation
`How many steps should we take?
`It is foolish to surmise that if we distributed enough
`
`14000
`
`12000
`
`10000
`
`Mean steps/day
`
`8000
`
`6000
`
`4000
`
`2000
`
`0
`
`1
`
`2
`
`4
`3
`Population groups
`
`5
`
`Legend
`1=8-10 year old children
`2=14-16 year old adolescents
`3= Healthy younger adults (approx. 20-50 years)
`4= Healthy older adults (>50 years)
`5= Individuals living with disabilities and chronic illnesses
`
`Figure 1. Expected values of steps/day for different populations
`
`Adapted in part from Tudor-Locke and Myers (2001b)
`
`4
`
`IPR2017-01058
`Garmin EX1012 Page 4
`
`
`
`pedometers to each household in the nation our work as
`physical activity promoters would be done. Without software
`(e.g., guidelines, programs, etc.) the hardware (i.e., the
`pedometer) is useless. Researchers and practitioners require
`practice guidelines (Tudor-Locke & Myers, 2001b),
`including step indices associated with important health-
`related outcomes (e.g., obesity, hypertension). As stated
`previously, Japanese health promotion efforts recommend a
`goal of 10,000 steps/day (Hatano, 1993; Yamanouchi et al.,
`1995). According to Hatano (1997) 10,000 steps/day is
`approximately equal to an energy expenditure of 300-400
`kcal/day (depending on walking speed and body size). This
`is double the amount (150 kcal/day) that the U.S. Surgeon
`General indicates is related to health benefits (U.S.
`Department of Health and Human Services, 1996).
`Compared to assembled expected values (Tudor-Locke &
`Myers, 2001b), 10,000 steps seems a reasonable estimate for
`ostensibly healthy adults, but there is currently little
`empirical evidence (i.e., linked to important health-related
`outcomes) to support such a threshold. Specifically, neither
`the appropriateness nor the sustainability of a universal goal
`of 10,000 steps/day has been thoroughly examined. Although
`body composition has been consistently related to
`pedometer-determined steps taken in adults (McClung et al.,
`2000; Tryon et al., 1992; C. Tudor-Locke et al., 2001; Tudor-
`Locke et al., 2002) and children (Rowlands, Eston, &
`Ingledew, 1999), few have attempted to link specific step cut
`points to indicators of body fatness. A single cross-sectional
`study has reported that individuals who take >9,000
`steps/day are more frequently classified as normal weight
`(defined by BMI cut points) and those who take <5,000
`steps/day are more frequently classified as obese; there were,
`however, exceptions to this rule (Tudor-Locke et al., 2001).
`Some experts have suggested that at least 15,000 steps/day is
`necessary to achieve weight loss goals (Leermakers, Dunn,
`& Blair, 2000).
`With regards to appropriateness as a universal goal, 10,000
`steps/day is likely too low for children; expected values for
`8-10 year old U.K. children currently range 12,000-16,000
`steps/day (Rowlands et al., 1999) while U.S. adolescents (14-
`16 years old) 600+ adolescents take 11,000-12,000 steps
`(Wilde, 2002). Further, a Japanese study (Suzuki et al., 1991)
`of handicapped youth ranging from 3 to 22 years reported
`that mentally retarded, blind, and deaf youth took an average
`of 14,500 steps/day, 12,700 steps/day, and 17,400 steps/day,
`respectively, compared to physically handicapped youth who
`took 8,050 steps/day. A recent study of 6-12 year old U.S.
`children reported 10,479-11,274 and 12,300-13,989
`steps/day for girls and boys respectively (Vincent &
`Pangrazi, in press). The 2001-2002 President’s Challenge
`Physical Activity and Fitness Awards Program (President’s
`Council on Physical Fitness and Sports, 2001) recommends
`that children accumulate 11,000 steps/day at least 5 days a
`week for a standard healthy base.
`With regards to sustainability, the 10,000 steps/day is
`unrealistically too high for sedentary individuals or those
`living with chronic diseases who take between 3,500-5,500
`steps/day (Tudor-Locke & Myers, 2001b). This would
`require a 2-3 fold increase in daily activity, setting up a high
`risk situation for failure and attrition. In support of this
`concern, a workplace walking program that prescribed
`
`10,000 steps/day has reported a high attrition rate (88% over
`12 weeks) (Iwane et al., 2000). In a study of healthy older
`adult exercisers (Tudor-Locke, Jones, et al., in press), half of
`the sample never achieved 10,000 steps on any single day of
`monitoring, despite the fact that they engaged in various
`forms of exercise (e.g., exercise class and informal walking
`for exercise) most days of the week. Wilde et al. (2001)
`reported that, even with a 30 minute walk included, the
`proportion of women who achieved >10,000 steps/day only
`ranged between 38-50%. Preliminary evidence suggests that
`the effort required to achieve a 10,000 step/goal is associated
`with reduced adherence in women participating in a
`pedometer-based intervention (Sidman, 2002b).
`Recommended levels of steps/day should not be determined
`from cross-sectional studies but should be inferred from
`longitudinal designs that allow step indices to emerge from
`the data related to important health outcomes. Until that
`time, a practical translation (in terms of pedometer-
`determined steps taken) of the public health recommendation
`would be useful to researchers interested in standardizing
`physical activity measures and practitioners charged with
`program evaluation and in motivating clientele to adopt
`healthful levels of physical activity. For example, it is
`possible to establish an index of steps taken in 30 minutes of
`brisk walking. Then we can compare this index to other
`activities of varying intensities and durations. We could also
`compare changes in steps/day due to intervention to this
`index to determine whether or not the public health
`recommendation was achieved. Welk et al. (2000) estimated
`that approximately 3,800-4,000 steps represented 30 minutes
`of moderate intensity walking. Wilde et al. (2001) reported
`that an unsupervised 30 minute walk included in a typical
`day of activity represented approximately 3,100 steps.
`Individuals with type 2 diabetes (mean age 53 years) took
`2,198±282 steps during a self-paced 20 minute walk,
`equivalent to 3,297 steps in 30 minutes (Tudor-Locke,
`Myers, Bell, Harris, & Rodger, in press). Measured directly,
`older (59-80 years) healthy individuals took 3,411±577 steps
`in 30 minutes of continuous walking (Tudor-Locke, Jones, et
`al., in press). Collectively, 3,100-4,000 pedometer-
`determined steps taken appear to be equivalent to 30 minutes
`of moderate intensity walking. A simple study is warranted
`to directly establish a reliable index of pedometer-determined
`steps equivalent to 30 minutes of moderate intensity activity.
`
`Another approach is to select personally relevant incremental
`goals anchored by individual baseline values. Any goal
`selected should be an improvement from baseline and should
`also be sustainable for the long term (Sidman, 2002a).
`Details are provided below on such an approach used in the
`First Step Program (Tudor-Locke, Myers, et al., in press;
`Tudor-Locke, Myers, & Rodger, 2000).
`
`How can pedometers be used to motivate and promote
`physical activity?
`A pedometer can be used as a tracking device (continuously
`collecting current activity), a feedback tool (providing
`immediate information on activity level), and as an
`environmental cue (reminder to be active). Used in
`combination with record keeping (e.g., calendars or diaries
`of daily progress), pedometers may be used in an effective
`
`5
`
`IPR2017-01058
`Garmin EX1012 Page 5
`
`
`
`way to increase daily physical activity. A number of
`implementation ideas have been suggested by Beighle et al.
`(2001) for use in school settings. Besides self-monitoring,
`however, a process of progressive goal-setting, reflection,
`and refinement should be put in place (Tudor-Locke, Myers,
`& Rodger, 2001).
`
`Base programs on theory
`An increasing number of intervention studies have used
`pedometers to track and/or motivate physical activity
`(Bassey, Patrick, Irving, Blecher, & Fentem, 1983;
`Fogelholm, Kukkonen-Harjula, & Oja, 1998; Iwane et al.,
`2000; Meshkinpour et al., 1998; Moreau et al., 2001; Puente-
`Maestu et al., 2000; Speck & Looney, 2001; Toda et al.,
`1998; Tudor-Locke, Myers, et al., in press; Yamanouchi et
`al., 1995). Project Active also used pedometers to increase
`lifestyle activity but did not provide details on pedometer use
`or outputs in the article (Dunn et al., 1999). With the
`exception of the First Step Program (Tudor-Locke, Myers, et
`al., in press; Tudor-Locke et al., 2000; Tudor-Locke et al.,
`2001), these studies provide few program details necessary to
`facilitate implementation and delivery in real-world settings.
`Program theory is used to systematically organize and
`explain what happens in the program and why (Myers, 1999;
`Sidani & Braden, 1998). Program theory drives program
`development, implementation, and evaluation. The essential
`elements of a program theory include: problem definition,
`critical inputs, mediating processes, expected outcomes,
`extraneous factors, and implementation issues (Lipsey, 1993;
`Sidani & Braden, 1998).
`
`Critical inputs are the key components or activities that must
`be present in successful physical activity interventions; these
`are informed by literature review, pilot work, consultation
`with program deliverers and recipients, and clinical
`understanding. Critical inputs in the First Step Program
`included initial group meetings with peers, flexibly
`scheduled moderate-intensity walking, self-monitoring and
`progressive ind