BioTrainer™ Clinical Studies

Introduction

According to the Centers for Disease Control and Prevention (CDC), obesity in the United States is epidemic and predictions are that 40% of Americans, or 68 million people, will be obese by the year 2010 (presented at the North American Association for the Study of Obesity, Oct, 2003). In a recent poll commissioned by the Harvard School of Public Health, obesity surpassed smoking as a public health concern (NY Times, Oct. 12, 2003). According to the CDC (2003), in the last 10 years, obesity rates have increased by more than 60% among adults and since 1980, obesity rates have doubled among children and tripled among adolescents.Despite the proven benefits of physical activity, more than 60% of American adults do not get enough physical activity to provide health benefits. This sedentary condition which few are addressing is “physical inactivity” and it’s seriously compromising our nation’s health.

As evidence is clear that physical inactivity is a contributing factor to the growing epidemic of obesity, there is a pressing need for new low-cost self-motivating tools to help both adults and children become more aware of, and change behaviors to improve their overall levels of exercise. New recording and assessment methods are needed to provide a means of accurately quantifying more complex components of physical activity and use this feedback to help motivate and promote increases in daily exercise.

Measuring Physical Activity:

Increasing evidence points to the need to know more about the health benefits of physical activity, and the effectiveness of activity interventions. A major limiting factor in studying and monitoring physical activity behaviors and the associated health benefits has been the lack of a reliable, valid, and standardized assessment tool (Melanson & Freedson, 1996). Currently, there is no universally accepted “gold standard” method for measuring physical activity (Melanson & Freedson, 1996). Assessing physical activity is a complex and challenging issue. There are a number of methods to assess physical activity and each has some relative advantages and disadvantages (Tudor-Locke & Myers, 2001). Self-report instruments are the most commonly used indicators of physical activity. Direct observation techniques have been commonly used to assess activity behavior. (McKenzie, et al, 2000, Elder, et al, 1998, Rowe, et al, 1997, Sallis and McKenzie, 1991). In most systems such as SOPLAY (McKenzie, et al, 2000) and SOFIT (Rowe, et al, 1997) an observer codes the type, and intensity of activity performed during a short periodic interval along with other details about the behavior or setting. The detail available through direct observation techniques offers some significant advantages for understanding activity behavior, but the time and cost of observations generally makes this type of assessment most practical for research or instructional applications. It is often used as a criterion measure to evaluate other methods.

The most effective tool for measuring physical exercise are various electronic motion sensing devices that capture details about activity patterns during normal daily life (Welk, Corbin, Dale, 2000). These instruments overcome the limitations of self-report and direct observation instruments and provide an objective indicator of physical activity. A formidable task for the use of these motion sensing devices is to provide an uncomplicated method of direct assessment and achievement scoring and reward that is easy to understand and administer.

A number of different electronic devices are available including heart rate monitors, pedometers and accelerometer-based activity monitors all of which have their own advantages and disadvantages. The sections below will provide more detail on these devices and lead to a justification for the measurement approach selected for the BioTrainer activity monitor.

Heart Rate Monitors:

Heart rate monitors are popular in many physical exercise programs to teach cardiovascular fitness and to track activity. These systems typically consist of a chest belt and integrated transmitter which sends a signal to a wrist-worn receiver. The wristwatch display usually provides a flashing heart icon for each detected heart beat and the rate is computed in beats-per-minute and displayed on a small LCD readout. While heart rate monitors can deliver useful information during specific exercise workouts they are not useful for tracking physical activity patterns under the normal lifestyle activities of daily living and rate can be influenced by nervousness, dehydration and stress (Welk, Corbin, & Dale, 2000b). They are prone to interference from other electrical equipment (and other monitors) and are inconvenient for participants to wear.

The inventors past experience with heart rate monitors have shown these devices to be seriously prone to frequent signal interruptions, radio frequency interference, and transmission problems. These limitations were experienced during extensive testing of the Polar OEM model #PCBA receiver and transmitter for use in our fitness equipment. When the transmitter belts were worn over extended periods of time and during more vigorous activities such as running or jumping, electrode contact became problematic and caused artifact-induced signal dropouts and rate errors. We also found the easy touch technology, where the finger is placed over a photo-sensor or across sensing pads on a wrist unit, to be difficult to manage requiring precise finger placement and little or no movement by the user to obtain an accurate reading.

While heart rate monitors can provide a useful indicator during specific bouts of exercise, they are not useful for tracking activity patterns under normal activities of daily living (Welk, Corbin, & Dale, 2000b). Welk’s research also cited another major problem with heart rate monitors is that the more physically active user may actually have a lower heart rate than their more sedentary counterparts, a variable, which could be confounding to the user. Further, heart rate does not typically increase or decrease to a large degree with normal life-style activities, although it does with more strenuous activities. Thus the heart-rate monitor is not sensitive enough to record the more subtle changes in physical activity which is easily accomplished with accelerometers.

Pedometers:

For a number of years, the pedometer (also known as a stepometer or step-counter) has been used successfully to motivate and assess physical activity and increase walking behavior (Todor-Lock, C., 2002). Although pedometers are very cost effective, one of the main flaws in using pedometers however is that they do not record intensity or velocity of movement nor do they have memory storage, restricting their use to measures of total accumulated steps per day. Pedometers are small step counting devices usually attached at the waist that count and display the number of steps performed during walking activity (Freedson and Miller, 2000). Some devices even contain AM/FM radios, provide verbal feedback, play music and pace your walking cadence. Pedometers can be found in almost any consumer catalog or retail store. Some of the more popular manufactures include Digi-Walker, Omron, Acumen, Bodytrend, Oregon Scientific, Sportline, Freestyle, Brookstone, AccuStep and many others. These clip-on devices are inexpensive, ranging from less than $15 up to $75 (Tudor-Locke & Myers, 2001a). Because the pedometer is relatively small and light weight (similar to the size of a pager), it is not intrusive (Welk et al., 2000a&b). In addition, virtually all segments of the population (including children, adults, and people with disabilities) can wear pedometers (Vincent & Pangrazi, 2002). Pedometers give people immediate feedback about how many steps they have taken during the day (Tudor-Locke, 2001b). This feedback can increase personal confidence and may also increase motivation to achieve a certain number of steps per day (Bassett, 2000). Moreover, a pedometer can be used as a coaching and self-monitoring tool (Freedson & Miller, 2000) to help people set goals (Bassett, 2000). A pedometer can also document change and recognize progress in daily activity levels (Beighle, Pangrazi, & Vincent, 2001). As a result, the pedometer has been a reasonably effective tool for helping people to increase physical activity levels.

The popularity of pedometers has increased in recent years, due in part, to the effective marketing and promotion of “10,000 step programs”. A study conducted at the Department of Medicine, Wakayama Medical College, Japan (Iwane, M., et al., 2000) investigated the effects of walking 10,000 steps per day or more (measured using a pedometer) on blood pressure and cardiac autonomic nerve activity in mild essential hypertensive patients. The results indicated that walking 10,000 steps per day or more was effective in lowering blood pressure, increasing exercise capacity, and reducing sympathetic activity in hypertensive patients. The idea of using 10,000 steps as a daily goal caught on quickly and a number of other step programs were born. A current Google search for 10,000 step programs draws 495,000 hits. Step counting programs, were initially introduced by companies that manufactured step-counting pedometers. The trend quickly spread to weight-loss programs, walking clubs, schools, state and local programs and to large national programs such as Queensland Health, Health Partners, Shape Up America, America on the Move and now the President’s Challenge Program has endorsed the use of pedometers as a method to increase physical activity. The American Diabetes Association also packages a step counter with its new book on the benefits of physical activity (Small Steps, Big Rewards) and McDonald's is offering a free counters with its new Happy Meal for adults, which is now being sold throughout their franchise.

While there are thousands of programs in place, there is no standardized model for setting baselines, realistic goals or for evaluating these programs. Most programs set their own guidelines and customize their own action plans but little, if any, is known about their effectiveness. There is also much confusion as to how many steps are optimal for age, weight, gender, body mass index and lifestyles. According to Greg Welk, a physical-activity researcher at Iowa State University and formerly with the Cooper Institute for Aerobics Research, "The number 10,000 has developed almost mythical proportions - it's actually not yet clear at what point you start getting a benefit" (Time, Oct. 13, 2003). Catrine Tudor-Locke, a health promotion expert at Arizona State University East suggested that “a common target is to achieve 10,000 steps per day. However this is likely to be too low a target for young people who could potentially achieve 12,000 to 16,000 but too high for the sedentary adult who may only reach 3,000 – 5,000. For the sedentary, setting a target of 10,000 steps presents a high-risk situation for failure and attrition. Goals set should be an improvement from baseline, realistic, achievable and sustainable over the long term” (Philadelphia Inquirer, Oct. 3, 2002). James Hill from the Center for Human Nutrition at the University of Colorado also emphasizes the use of a baseline approach and suggests that 2,000 steps above baseline is a modest goal that will burn roughly 100 calories (Associated Press, July 14, 2003).

Limitations of Pedometers:

Despite the appeal and popularity of pedometers, they also have many mechanical and functional limitations. Most pedometers use a simple spring suspended lever arm (sensor) that moves up and down and makes contact with ambulation. An electrical circuit closes with each step and the accumulated step count is displayed digitally on a LCD screen. Some devices require the user to set the tension or sensitivity of the sensor arm for accurate counting. Improper adjustment will seriously affect the accuracy of the device (IM Systems engineering evaluation) producing significant over-counts or under-counts. The main problem however is that they do not measure the intensity (how hard), duration (how long), or frequency (how often) physical activity occurs (Beighle, Pangrazi, & Vincent, 2001). In addition, pedometers generally underestimate the number of steps taken during higher intensity activities (Rowlands, Eston, & Ingledew, 1999) and show consistently more errors during slow walking (Bassett et al., 1996). Discrepancies among different models of pedometers can also limit their usefulness (Freedson & Miller, 2000). Furthermore, pedometers display steps taken over a period of time and are not able to store or recall single-day values over a period of many days. Thus, daily written records must be kept if the pedometer is reset daily. Pedometers are also not as accurate for people who do a fair amount of bending and/or who have excessive abdominal fat, as the pedometer may move away from the person’s body (Tudor-Locke & Myers, 2001b). Additionally, pedometers cannot distinguish between walking and running (Bassett, 2000) and cannot measure static movements, non-locomotor activities, upper-body exercises and cycling (Beighle, Pangrazi, & Vincent, 2001).

Activity Monitors (BioTrainer product):

Recent studies and technological advances have shown that frequency, duration and intensity of physical activity can be objectively measured by wearable acceleration-based activity monitors, and these devices can accurately record body movements. Notwithstanding their higher cost, accelerometer measuring techniques have been shown to provide a high level of accuracy for objectively assessing physical activity (Fairweather, et al., 1999; Ott, et al., Welk, et al., 1999, 2000b; Trost, et al., 2000; Nilsson, et al., 2002; and Sallis, et al., 2002). A variety of commercially available activity monitors are available to measure complex physical activity patterns, display and store the data and some designs provide download for PC analysis. These devices are usually small and unobtrusive but differ from pedometers in that they can sample and store detailed information about intensity levels, movements and physical activity patterns (Kohl et al., 2000). Ranging in price to well over $2,000, these monitors come in many forms and are designed to perform a variety of functions but have been largely used for research applications. For instance, IM Systems manufactures 5 commercial models: The “BioTrainer-II” for physical activity and fitness (patent #5,749,372), the “BioTrainer-Pro” for sports medicine and rehabilitation, the “ActiTrac” for sleep disorders and behavioral assessment (FDA Cleared # K992410), the “PAM-RL” for restless legs and period limb movements (FDA Cleared # K010997) and the “DigiTrac” for Parkinson’s tremor and dyskinesias.

The main component of an accelerometer monitor is the sophisticated sensor assembly, which is generally made of piezoelectric, piezoceramic or piezofilm materials and can be broadly termed as either uniaxial or triaxial. Uniaxial accelerometers measure accelerations in a single plane, and can be attached to the trunk or limbs, whereas triaxial accelerometers measure accelerations along three planes; vertical, anterior-posterior, and medio-lateral (Freedson and Miller, 2000). The principal function of accelerometers is that the sensor converts physical movements into electrical signals that are proportional to the muscular force producing motion (Melanson and Freedson, 1996). The signals are then sampled by an analog to digital converter (usually 10 to 40 Hz.), and input to a microprocessor for storage and consequent download for analysis. Some activity meters also provide for a direct panel readout and/or graph of activity measures on a LCD display such as IM Systems’ BioTrainer model. Accelerometer monitors can store raw digitized movement data directly into memory, or average the data for storage in selected epochs of time (e.g. 2, 15, 30, 60 or 120 seconds). They can also be used to estimate caloric burn and energy expenditure (Welk, et al.,2000). The monitors are small, easy to use and well suited to assessing physical activity in both children and adults.

In addition to these attributes, the overwhelming advantage of newer generation accelerometers is that the embedded microprocessor can be custom programmed to provide any number of direct display features and functions for supporting simple or complex behavioral interventions and protocols.

BioTrainer Activity Counts:

Physical activity recorded by an accelerometer-based activity monitor is derived from the output of an accelerometer sensor using standard analog to digital conversion (A/D) methods and digitally sampled at a rate of 40 samples per second with a resolution of 10 bits or 1024 parts per sampled value. These values are then averaged by means of digital integration and converted into activity counts or for display on the LCD readout. This sensitivity range will capture activity from subtle movements (about 0.05-0.2 g) to moderate activity (about 0.2 to 1.0 g) to vigorous walking and jogging (about 1.0-4.0 g). These sensitivity ranges are approximate for varied populations and were determined by clinical investigations (IM Systems) and through an NIH-SBIR grant #5R44MH52049 awarded to IM Systems which evaluated the movement and acceleration characteristics of participants in a treadmill/weight loss study described below (APA, 1998). Validated formulations for quantifying average body acceleration, developed for IM Systems’ BioTrainer activity monitors, are programmed into the monitor’s firmware and processed in real-time for direct display of Activity Counts. This conversion (a derivative of the method described in IM Systems’ U.S. Patent # 5,749,372) uses the formula where 1.0 “g” of motion (referenced to earth gravitation) equals 9.807 m/sec2. The equation gave = 3.071 x [Activity Count / time period (seconds)] are used to convert acceleration of body motions into activity points. The constant of 3.071 is derived as follows: the Actometer will be computer calibrated to accumulate an activity count of 1.0 for a period of 32.2 seconds when placed on a moving sinusoidal calibration table set to 0.3 g peak-to-peak at a rate of 4 hertz. The activity rate is therefore equal to 1.0 / 32.2 seconds which equal 0.03106 activity units/second. Average acceleration is = 0.3 g’s p-p x (0.636 / 2) = 0.0954 gave. Therefore: gave/activity rate = 0.0954 / 0.03106 = 3.071. To increase the resolution or adjust the activity count scale, the constant can be divided by a fixed value (e.g. 3.017/20 will increase the Activity Counts by a factor of 20).

Also, a recent study (Welk, Almeida, Morss, 2003) supported the validity of the BioTrainer acceleration technology (programmed with above algorithm) for assessing physical activity.

BioTrainer Calorie Burn Counter:

The BioTrainer measures horizontal and vertical body acceleration. When positioned (clipped) to the front part of a person’s waist, this acceleration is correlated to exercise calories during walking, jogging or running activities. Additionally, the BioTrainer adjusts exercise calories accumulated in proportion to a person’s body weight selected.

The BioTrainer correlation technique is based on the fourth edition of Exercise Physiology, Energy, Nutrition, and Human Performance [1]. Table 10.1 of this publication gives predictions of energy expenditure versus walking speed and body weight. From the values of Table 10.1, treadmill walking tests at IM Systems and the VO2 oxygen uptake studies performed at the Cooper Institute of Aerobics Research, the following equation was derived that approximates exercise calories versus velocity and body weight.

Equation 1. C = (.00719W + .32564) V

C = calories per minute

W = body weight (lbs)

V = walking, jogging, running velocity (mph)

The BioTrainer measures the average body horizontal and vertical acceleration when clipped to the front part of a person’s waist. This acceleration is directly proportional to body velocity as indicated by the following equation.

Equation 2. A = K ˜ V

A = average positive acceleration per minute

K = BioTrainer scale factor

V = body velocity (mph)

From equation 2, V = A / K

Substituting for V in equation 1 yields,

Equation 3. C = (.00719W + .32564) ˜ A / K

This equation is used in the BioTrainer monitor to correlate body acceleration and body weight to exercise calories expended.

Activity Count Study:

In a weight loss study (publish in JAMA by Anderson, R., et al, 1999), the BioTrainer-Pro monitor, programmed to display activity points, was given to 20 overweight female subjects instructed to increase their lifestyle activities by wearing the monitor with visual feedback. A second group of 20 overweight female subjects attended 3 aerobic classes per week. After 16 weeks, the study showed that increased casual lifestyle activity was just as effective as exercise classes for weight loss and both groups lost about 17.5 pounds. But after 12 months, the group using the BioTrainer gained only 0.5 pounds while the aerobics group regained 3.5 pounds. This study demonstrated both that increased lifestyle activity may be as effective as structured exercise classes and that visual activity feedback counts was motivational and helped reduce the “yo-yo” effect during weight-loss maintenance. (Research was supported by National Research Service Award #DK09241 and the National Institutes of Health - NIH #MH0070).

Activity Count Calibration and Test:

Piezoelectric accelerometer technology is well developed at IM Systems and used in most all of the company’s ambulatory activity monitors. Existing methods of absolute g-force calibration including a variable acceleration table is use to set range and sensitivity of the movement detection circuits. Bench testing and calibration is performed at IM Systems’ development laboratory on each monitor to comply with established standards. The analog and data measurements must all match within acceptable tolerance of no more than ± 0.5% @ 3Hz. Tolerance for acceleration frequencies above or below 3Hz but must fall within ±1.0% for the full specified bandwidth of the monitor (0.1 to 12 Hz.).

The BioTrainer is currently tested and calibrated using the same set-up and methods as for our BioTrainer monitors. Calibration and test is accomplished by placing the monitor on a moving platform fixture that oscillates up and down in the vertical axis. The calibration consists of a long throw 15" speaker with cone modified to accept an oscillating platform table. A low frequency signal adjustable from 0.1 Hz to 12 Hz will drive the platform with a sinusoidal motion. Exact calibration is achieved by also mounting on the test platform a precision accelerometer (IC Sensor type 3145) as the standard with its signal output directed to a lab PC programmed with calibration software written in Visual Basic-6. Acceleration sensitivity is adjusted through an automated method that inserts a calibration factor into the ROM memory of the Actometer’s microprocessor to bring each unit to within the specified range. This software calibration method therefore automatically compensates for the accelerometer sensors or amplifiers that are naturally outside the prescribed tolerances. An automatic calibration procedure for high production is now being considered. If no calibration procedures are employed, variations of about 15% max are typical. This is well within the accuracy limit for a consumer product.

Literature Cited:

Allensworth D., Lawson, E., Nicholson, L., & Wyche, J. (Eds.). (1997). Schools and Health: Our Nation’s Investment. Washington, DC: National Academy Press.

Andersen, R.E., Wadden, T.A., Bartlett, S.J., Zemel, B., Verde, T.J. & Franckowiak, S.C. (1999). Effects of lifestyle activity vs. structured aerobic exercise in obese women: A randomized trial. Journal of the American Medical Association, 281(4), 335-340.

Bassett, D. R. (2000). Validity and reliability issues in objective monitoring of physical activity. Res. Q. Exerc. Sport 71(2): 30-36.

Bassett, D. R., Ainsworth, B.E., Leggett, S.R., Mathien, C.A., Main, J.A., Hunter, D.C., and Duncan, G.E. (1996). Accuracy of five electronic pedometers for measuring distance walked. Med. Sci. Sports Exerc. 28(8): 1071-1077.

Beighle, A., Pangrazi, R. P., & Vincent, S. D. (2001). Pedometers, physical activity, and accountability. JOPERD, 72 (9), 16–19.

Brownell (Eds.), Eating disorders and obesity: A comprehensive handbook (2nd ed., pp. 131–135). New York: The Guilford Press.

Differding, J.A., Welk, G.J. & Hyman, A.S. (1999). Physical activity patterns among typically sedentary adults. Medicine and Science in Sports and Exercise, 31, 5 (supplement), S164.

Elder, J.P., Broyles, S.L., McKenzie, T.L., Sallis, J.F., Berry, C.C., Davis, T.B., Hoy, P.L., & Nader, P.R. (1998). Direct home observation of the prompting of physical activity in sedentary and active Mexican and Anglo-American Children. Journal od Developmental and Behavioral Pediatrics, 19, 26-30.

Fairweather, S. C., Reilly, J.J., Grant, S., Whittaker, A., and Paton, J.Y (1999). Using the Computer Science and Applications (CSA) activity monitor in preschool children. Pediatr. Exerc Sci. 11: 413-420.

Freedson, P. S., and Miller, K. (2000). Objective monitoring of physical activity using motion sensors and heart rate. Res. Q. Exerc. Sport 71(2): 21-29.

Gorny, S.W. & Allen, R.P. (1999). What is an activity count?: A comparison of different methodologies used in wrist actigraphy. Sleep, 22(supplement), S52.

Gorny, S.W., Allen, R.P., Krausman, D.T., Cammarata, J. & Earley, C.J. (1997). A parametric and sleep hysteresis approach to assessing sleep and wake from a wrist activity meter with enhanced frequency range. Sleep Research, 26, 662.

Gorny, S.W., Allen, R.P., Krausman, D.T. & Earley, C.J. (1996). Parametric analyses of factors affecting accuracy for detection of wake epochs after sleep onset based on wrist activity data. Sleep Research, 25, 490.

Heshka,S., Anderson, J., Atkinson, R., Greenway, F., Hill, J., Phinney, S., Kolotkin, R., Miller-Kovach, K., Pi-Sunyer, X. (2003). Weight Loss With Self-help Compared With a Structured Commercial Program . JAMA, 289:1792-1798.

Iwane, M., Arita, M., Tomimoto, S., Satani, O., Matsumoto, M., Miyashita, K., & Nishio, I. (2000). Walking 10,000 steps a day or more reduces blood pressure and sympathetic nerve activity in mild essential hypertension. Hypertension Research, 23, 573-580.

Jones, S.L., Wood, K., Thompson, R. & Welk, G.J. (1999). Effect of monitor placement on output from three different accelerometers. Medicine and Science in Sports and Exercise, 31, 5 (supplement), S142.

Kiningham, R. B. (2001). Exercise and primary prevention of cardiovascular disease. Clinics in Family Practice, 3, 707–732.

Kohl, H. W., Fulton, J.E., and Casperen, C.J. (2000). Assessment of physical activity among children and adolescents: A review and synthesis. Prev. Med. 31: S54-S76.

Kripke D, Mullaney D, Messin S, Wyborney J. Wrist actigraphic measures of sleep and rhythms. Electroencephalogr Clin Neurophysiol. 1978;44:674-676.

Lamonte, M. J., & Ainsworth, B. E. (2001). Quantifying energy expenditure and physical activity in the context of dose response. Medicine and Science in Sports and Exercise, 33, S370–S378.

Leidy NK, Abbot RD, Fedenko KM. Sensitivity and reproducibility of the dual-mode actigraphy under controlled levels of activity intensity. Nurs Res. 1997;46:5-11.

Mckenzie, T.L., Marshall, S.J., Sallis, J.F., & Conway, T.L. (2000), Leisure-time physical activity in school enviroments: An observational study using SOPLAY. Preventive Medicine, 30, 70-77.

McKenzie, T. L. (1991). Observational measures of children's physical activity. Journal of School Health, 61, 224-227.

Melanson, E. L., Jr., & Freedson, P. S. (1996). Physical activity assessment: A review of methods. Critical Reviews in Food Science and Nutrition, 36, 385–396.

Nilsson, A., Ekelund, U., Yngve, A., and Sjostrom, M. (2002). Assessing physical activity using different time sampling intervals and placements. Pediatr. Exerc Sci. 14: 87-96.

Ott, A. E., Pate, R.R., Trost, S.G., Ward, D.S., and Saunders, R. (2000). The use of uniaxial and triaxial accelerometers to measure chidren's "free play" physical activity. Pediatr. Exerc Sci. 12: 360-370.

Redmond DP, Hegge FW. Observations on the design and specification of a wrist-worn human activity monitoring system. Behavior Research Methods, Instruments and Computers. 1985;17:659-669.

Rowe, P.J., Schuldheisz, J.M., & van der Mars, H. (1997). Measuring physical activity in physical education: Validation of the SOFIT direct observation Instrument for use with first to eighth grade students. Pediatric Exercise Science. 9(2), 136-149.

Rowlands, A. V., Eston, R. G., & Ingledew, D. K. (1999). Relationship between activity levels, aerobic fitness, and body fat in 8- to 10-yr-old children. Journal of Applied Physiology, 86 (4), 1428–1435.

Safer, D.J. & Allen, R.P. (1998). Feedback of activity in the treatment of obesity. Paper presented at the Annual Meeting of the American Psychiatric Association, May 30 - June 4, 1998, Toronto, Canada.

Sallis, J. F., Taylor, W.C., Dowda, M., Freedson, P.S., and Pate, R.R. (2002). Correlates of vigorous physical activity for children in grades 1 through 12: Comparing parent-reported and objectively measured physical activity. Pediatr. Exerc Sci. 14: 30-44.

Sallis, J.F., McKenzie, T.L., Elder, J.L., Galati, T., Berry, C.C., Zive, M., Nader, P.R. (1998). Sex and ethnic differences in children’s physical activity: Discrepancies between self-report and objective measures. Pediatric Exercise Science, 10, 277-284.

Trost, S. G., Pate, R.R., Freedson, P.S., Sallis, J.F., and Taylor, W.C. (2000). Using objective physical activity measures with youth: How many days of Monitoring are needed? Med. Sci. Sports Exerc. 32(2): 426-431.

Tudor-Locke, C. (2002). Taking steps toward increased physical activity: using pedometers to measure and motivate. Presidents Council on Physical Fitness and Sports Research Digest, 3(17).

Tudor-Locke, C. E., & Myers, A. M. (2001a). Challenges and opportunities for measuring physical activity in sedentary adults. Sports Medicine, 31, 91–100.

Tudor-Locke, C. E., & Myers, A. M. (2001b). Methodological considerations for researchers and practitioners using pedometers to measure physical (ambulatory) activity. Research Quarterly for Exercise and Sport, 72 (1), 1–12.

Tudor-Locke (2001c). A preliminary study to determine instrument responsiveness to change with a walking program: Physical activity logs and pedometers. Res. Q. Exerc. Sport 72(3): 288-293.

Vincent, S. D., & Pangrazi, R. P. (2002). Does reactivity exist in children when measuring activity levels with pedometers? Pediatric Exercise Science, 14, 56–63.

Welk. G., Almeida, J., Morss, G., (2003). Laboratory calibration and validation of the Biotrainer and Actitrac activity monitors. Med Sci Sports Exerc. 2003 Jun;35(6):1057-64.

Welk, G.J., Blair, S.N., Wood, K.W., Jones. S., and Thompson, R.W. (2000a). A comparative evaluation of three accelerometry-based activity monitors. Medicine and Science in Sports and Exercise, 32(9), S489-S497.

Welk, G. J., and Corbin, C.B., and Dale, D. (2000b). “Measurement issues in the assessment of physical activity in children.” Res. Q. Exerc. Sport 71(2): 59-73.

Welk, G., Differding, J. A., Thompson, R. W., Blair, S. N., Dziura, J., & Hart, P. (2000c). The utility of the digi-walker step counter to assess daily physical activity patterns. Medicine and Science in Sport and Exercise, 32 (9), S481–S488.

Welk, G.J., Wood, K., Jones, S.L. & Barlow, C.E. (1999). The validity of three different accelerometers for the assessment of lifestyle physical activity. Medicine and Science in Sports and Exercise, 31, 5 (supplement), S142.

Wood, K. & Welk, G.J. (1999). Variability in activity patterns in children. Medicine and Science in Sports and Exercise, 31, 5 (supplement), S321.