Share this article on:

00005768-199804000-0002400005768_1998_30_634_jakicic_relationship_4article< 63_0_6_5 >Medicine & Science in Sports & Exercise© Williams & Wilkins 1998. All Rights Reserved.Volume 30(4)April 1998pp 634-638Accuracy of self-reported exercise and the relationship with weight loss in overweight women[Special Communications: Methods]JAKICIC, JOHN M.; POLLEY, BETSY A.; WING, RENA R.University of Pittsburgh School of Medicine, Pittsburgh PA 15213Submitted for publication March 1996.Accepted for publication June 1997.ABSTRACTPurpose: The purpose of this study was to compare self-reported exercise to a more objective measurement of exercise (i.e., Tri-Trac Accelerometer) and to assess whether there is a difference in weight loss between individuals who under- and over-report their exercise.Methods: Fifty overweight females (BMI = 34.0 ± 4.2 kg·m-2) who were participating in a behavioral weight control program were included in this study. Subjects were randomly assigned to a long-bout or a short-bout exercise condition, with both groups instructed to exercise 30 min·day-1 on 5 d·wk-1 for a period of 20 wk. The long-bout group was to exercise in one continuous session (e.g., one 30-min session per day), whereas the short-bout group was to divide the exercise into multiple 10-min sessions (e.g., three 10-min sessions per day). Subjects recorded their exercise in a daily exercise log and wore a Tri-Trac accelerometer for a 1-wk period to validate self-reported exercise bouts.Results: Results showed that approximately 45% of the women over-reported the amount of exercise that they performed, and this did not differ between the long-bout and short-bout groups. Women who over-reported their exercise had significantly poorer weight loss across the 20-wk program than women who under-reported their exercise (6.3 ± 3.6 kg vs 9.4 ± 5.2 kg).Conclusions: The results of this study suggest that overweight women who over-report their exercise will have poorer weight loss while enrolled in a behavioral weight loss program compared with others enrolled in the program, and the Tri-Trac Accelerometer may be useful in identifying individuals who inaccurately report the amount of their exercise. The ability to classify individuals as either over- or under-reporters of their exercise may be helpful to weight loss therapists and lead to more successful treatment for obesity.Exercise is a key component in behavioral weight control programs(2,3,9). Despite the known importance of exercise in a weight loss program, it has been difficult for researchers and clinicians to quantify accurately the amount of exercise that patients are doing and to determine whether patients are adhering to the exercise prescription. Many studies that have examined exercise patterns have relied on self-reported data provided by the subjects in the form of an activity log or activity questionnaire. However, concern has been raised about the accuracy of such self-reported exercise data, and this may be especially true for overweight individuals (6). This inaccurate reporting of exercise may lead to an energy imbalance which may be reflected in one's success in a weight loss program. For example, an individual may report 300 calories of activity per day, but in actuality is only expending 150 calories per day. Thus, this individual will lose less weight over time than would be expected based on self-reported exercise.A newly developed tool for measuring physical activity is the Tri-Trac accelerometer (Hemokinetics, Madison, WI), a device that measures vertical, horizontal, and lateral motion. Studies have shown a strong correlation between tri-axial accelerometer output and the metabolic cost of physical activity (1). Further, Matthews and Freedson(7) showed a significant correlation between total daily energy expenditure recorded by the Tri-Trac and daily energy expenditure estimated from both 3-d activity logs and a 7-d physical activity recall. In addition, the Tri-Trac can store minute-by-minute data for up to 14 consecutive days, and this feature allows investigators to more closely examine exercise patterns of patients in a free-living environment(10). This technology provides an objective method to assess exercise patterns (i.e., the amount of exercise that patients participate in and when this exercise is actually being done) and thus can be used to verify self-reported exercise records. We have previously used the Tri-Trac Accelerometer to examine exercise patterns in obese subjects instructed to exercise in either a continuous (i.e., 40 min) or discontinuous(i.e., multiple 10-min sessions) manner (4). The analysis of the group Tri-Trac data showed that the duration of the exercise sessions was similar to what was prescribed. However, no comparison of Tri-Trac and self-reported exercise data was done for each individual subject to examine whether this information could be used to identify specific individuals who inaccurately report their exercise behaviors and the impact that this could have on weight loss. By knowing this information, a weight loss therapist may better be able to individualize treatment in overweight patients that are having difficulty losing weight.This study, which is an extension of previously published data described above (4), used the Tri-Trac Accelerometer during 1 wk of a behavioral weight loss program to determine: 1) whether there are overweight individuals who either over-report or under-report their exercise participation, and 2) whether there are differences in weight loss during a 20-wk behavioral weight loss program between overweight individuals who either under-report or over-report their exercise.METHODSSubjects. Fifty overweight females participating in a behavioral weight control program were included in this study. Subjects were recruited from advertisements placed in local newspapers seeking participants for a weight loss program. The age of the subjects ranged from 25 to 50 yr, with body weight ranging from 120% to 175% of ideal body weight according to the Metropolitan Life Insurance Company Tables (8). Patients were excluded if they met any of the following criteria: 1) a history of myocardial infarction or stroke, 2) a history of diabetes, 3) orthopedic problems that would limit participation in an exercise program, 4) metabolic disorders (e.g., diabetes mellitus), 5) were taking prescription medication that would affect heart rate or blood pressure (e.g., beta blockade), 6) were pregnant or had been pregnant within the previous 6 months, 7) were participating in a regular exercise program (at least 3 d·wk-1 for 20 min·d-1), 8) were participating in another weight loss program. The study protocol was approved by the Institutional Review Board at the University of Pittsburgh and written informed consent was provided by all subjects before participation. Patients were not compensated for their participation in this study, nor were they compensated for being compliant to the dietary and/or exercise intervention. Characteristics of these subjects are shown in Table 1.TABLE 1. Descriptive baseline characteristics of subjects (mean ± SD).Treatment. All subjects in this study were participants in a treatment study investigating the effects of prescribing exercise in long versus short bouts on exercise adherence. Subjects were randomly assigned to either a long-bout (LB) exercise group which was instructed to exercise 1 time per day for a period progressing from 20 to 40 min in duration, or to a short-bout (SB) exercise group which was instructed to exercise in bouts of 10 min, and progressed from 2 to 4 bouts per day. Thus, the total amount of time prescribed was identical in the LB and SB conditions, but the LB group performed 1 bout per day whereas the SB group performed multiple bouts of 10 min each day. All subjects were also instructed to eat 1200-1500 kcal·d-1 and to reduce fat intake to 20% of total calories. The components of these interventions are shown in Table 2 and have been described in detail (4).TABLE 2. Description of the intervention components.Self-reported physical activity and dietary intake. Subjects were instructed to keep a record of the exercise they performed for each week of the program. These exercise records included the day and the time of day the exercise was performed, the type of exercise, and the duration of the exercise session. Subjects were instructed to record only the time that they were active, excluding the time spent stretching, etc. before and after the exercise sessions. Subjects were also instructed to record their food intake and were taught how to compute the calorie and fat content of these foods. These completed exercise records were collected weekly by the principal investigator, and the data (number of exercise sessions, exercise time) were entered into computerized database. Compliance to the dietary recommendations was computed as a percentage of the number of days the self-reported calorie intake did not exceed the prescribed level (i.e., 1200-1500 kcal·d-1).Tri-Trac Accelerometer. A Tri-Trac Accelerometer (Hemokinetics, Madison, WI) was used to provide an objective measure of physical activity. The Tri-Trac Accelerometer provides information about energy expenditure for each minute of the day, with physical activity taken into account by measuring acceleration resulting from horizontal, vertical, or lateral motion. Using regression equations provided by the manufacturer, the acceleration data and subject characteristics (height, weight, age, gender) are used to estimate energy expenditure at 1-min intervals. The Tri-Trac can collect minute-by-minute data for a period of 14 d, and this data can be retrieved by a microcomputer using software provided by the manufacturer.Only ten Tri-Trac Accelerometers were available to be used by the 50 subjects in this study. Because of this limited number, subjects were randomly assigned to wear a Tri-Trac for a 7-d period between weeks 5 and 10 of the treatment phase of this study. Because the subjects received the Tri-Trac on the afternoon of day 1, only the 6 full days (days 2-7) were used for comparison with self-reported exercise records.Data from the Tri-Trac were examined to identify periods of time when the subject was expending at least 3 kcal·min-1. Based on the characteristics of the subjects in this study, 3 kcal·min-1 would be equivalent to 2 metabolic equivalents (METs). Therefore, this would allow us to identify periods of time when an individual was active compared with when he/she was sedentary. A computer analysis program, developed in our laboratory using SAS, provided information about the time of day that an activity ≥ 3 kcal·min-1 began and the number of consecutive minutes that the activity was sustained at an energy expenditure of ≥ 3 kcal·min-1. For example, if energy expenditure was ≥ 3 kcal·min-1 at 10:00 a.m. and remained at this level until 10:15 a.m., the computer program would report a duration of 15 min in which energy expenditure ≥ 3 kcal·min-1 was sustained. Exercise sessions recorded by the Tri-Trac that were ≥ 20 min in the LB group and ≥ 10 min in the SB group (and hence fulfilled the exercise prescriptions for the two groups, respectively) were identified and used for analysis.Because the Tri-Trac can provide the actual time of day when an exercise bout occurred, we also attempted to match the self-reported exercise sessions with the sessions recorded by the Tri-Trac. A match was defined as an exercise bout recorded by the Tri-Trac that occurred within 60 min of the time of day that the subject reported doing the exercise according to their self-reported record. For example, if an individual reported starting an exercise bout at 10:00 a.m. and we found an exercise bout on the Tri-Trac that started at 10:30 a.m., this would be considered a match.Body Weight. Body weight was assessed at 0 and 20 wk of a behavioral weight control program using a calibrated medical balance beam scale (Detecto, Webb City, MO). In addition, body weight was assessed the day that the Tri-Trac was given to the participant so that it could be programmed accurately before collecting energy expenditure data.Statistical Analysis. Statistical analyses were performed using Statistical Packages for the Social Sciences (SPSS for Windows, Version 6.0.1) software. Chi-square was used to analyze whether there was a difference in the proportion of individuals who under-reported or over-reported their exercise. In addition, independent t-test were used to assess whether the changes in body weight across the 20-wk treatment program differed between those individuals categorized as over- or under-reporting their exercise. Statistical significance was defined as P < 0.05.RESULTSThe first purpose of this study was to identify individuals who either over-reported or under/accurately reported their exercise during the observation week. For these analyses we examined the ability of subjects to report accurately both the number of exercise bouts performed and the amount of time spent exercising. In the LB group, 54.2% under-reported and 45.8% over-reported the number of exercise bouts they had over the 6-d period. In the SB group, 61.5% of the subjects under-reported and 38.5% over-reported the number of exercise bouts. A chi-square to test these proportions was not significant, and these results are shown in Table 3. A similar analysis done on the amount of exercise time (rather than the number of exercise bouts) showed no difference in the proportion of LB and SB subjects who over- or under-reported their exercise time (seeTable 3). Therefore, because there was no difference in the proportion of subjects in the LB and SB groups who under- and over-reported their exercise, these groups were combined for further analyses. The differences between the self-reported exercise and exercise recorded by the Tri-Trac Accelerometer are presented in Table 4.TABLE 3. Chi-square to assess the distribution of the percentage of subjects in the long-bout and short-bout exercise groups that under-reported and over-reported their exercise compared to Tri-Trac Accelerometer.TABLE 4. Self-reported exercise, exercise recorded by the Tri-Trac Accelerometer, and weight loss of subjects that under-reported and over-reported the minutes of exercise and number of exercise sessions that they perform (mean ± SD.)Analyses were performed to examine whether the amount of self-reported exercise during the week that the Tri-Trac was worn differed from the amount of self-reported exercise during the weeks in which the exercise prescription was identical but the Tri-Trac was not worn. For example, if the Tri-Trac was worn during week 6, this week was compared to weeks 5-8 because the exercise prescription was the same (see Table 2). Results showed that wearing the Tri-Trac did not increase the amount of self-reported exercise (data not shown).When we attempted to actually match a self-reported bout of exercise with an exercise bout occurring at a similar time of the day from the Tri-Trac, we found that 88.5 ± 24.2% of the self-reported sessions were matched with an exercise session from the Tri-Trac in individuals who were classified as under-reporting their exercise sessions (see above), whereas 44.0 ± 28.1% of the self-reported sessions were matched with a session from the Tri-Trac for individuals who were classified as over-reporting their exercise sessions. This difference between under- and over-reporters was statistically significant (P < 0.001). For the matched sessions individuals classified as under-reporters actually under-reported their total exercise time by 13.1 ± 25.8 min, whereas those classified as over-reporters actually over-reported their exercise time by a total of 1.3 ± 17.2 min(P < 0.03). We repeated these analyses to compare individuals classified as under- and over-reporters according to exercise time (see above), and the pattern of results was similar (data not shown).The second purpose of this study was to determine whether there were differences in weight loss during the treatment program when subjects were grouped according to the types of errors made in reporting exercise (e.g., under-versus over-reporting of exercise duration or number of exercise bouts). Weight loss data was available for 48 of the 50 subjects, and only these subjects were included in the following analyses. Differences in weight loss between subjects who under/accurately reported and over-reported their exercise bouts were analyzed using an independent t-test. Overall, subjects who under/accurately reported the number of exercise bouts lost 9.4± 5.2 kg, whereas subjects who over-reported the number of exercise bouts lost 6.3 ± 3.6 kg (P < 0.02). A power analysis was performed based on the sample size used in this study, and the results showed that the power to detect this difference in weight loss was 0.64. The weight loss of subjects who under-reported the amount of time they spent exercising was 9.3 ± 5.2 kg compared to 6.8 ± 4.1 kg for those who over-reported their exercise time (P < 0.07). The weight loss data are presented in Table 4.These differences in weight loss between subjects who over-reported versus those who under/accurately reported their exercise might also reflect a difference in adherence to the prescribed dietary intervention. To assess this, the dietary intake records (described in the methods section) were analyzed to compare differences in number of days that the subjects were compliant to either the calorie or fat daily intake goals (seeTable 5). No significant differences were found between subjects who over-reported their exercise versus those who under-reported their exercise for either of these variables. Further, no differences were found when the data were analyzed separately for the LB and SB groups.TABLE 5. Compliance to dietary component of the intervention by subjects that under-reported and over-reported their exercise.DISCUSSIONSome evidence supports the belief that overweight individuals tend to over-estimate or over-report the amount of activity they perform(6). The study conducted by Lichtman et al.(6) observed participants for a 14-d period and compared the total self-reported (at 15-min intervals) physical activity with total energy expenditure measured using doubly labeled water. The current study was different from previous studies in this area in that we attempted to confirm individual exercise bouts rather than total energy expenditure. Examination of the data revealed a group of women who over-estimated and a group of women who underestimated the number of exercise bouts they performed or the amount of time that they spent exercising (see Table 3). Results of chi-square analyses suggest that the proportion of subjects that over- or under-reported their exercise was unaffected by the way in which the exercise was prescribed (i.e., LB vs SB groups). However, there was a difference in weight loss, with subjects who over-reported their exercise losing less weight than those subjects who under-reported their exercise. Therefore, using an objective measure to identify individuals who over-report their exercise may be useful in identifying those individuals who will also be less successful in a behavioral weight loss program.The results of this study may have practical application to weight loss therapists. One reason that individuals may over-report their exercise could be that they lack the skills to accurately self monitor. In this situation, further training of the individual on self-monitoring techniques may be sufficient to correct this problem. However, individuals may also over-report their exercise in an attempt to mask their inability to achieve the exercise goal. In this case, by using the Tri-Trac to identify individuals who over-report their exercise, a therapist may be able to implement strategies that facilitate the participant's achievement of the exercise goal. This may include such behavioral strategies as problem solving, or it may require a more realistic exercise goal to be established for this individual. A third scenario is that a participant is accurately reporting exercise and is achieving the recommended exercise goal but is failing to lose weight. Under these circumstances, the therapist can assume that the participant may be having difficulty with the dietary component of the program rather than the exercise component and can focus on using intervention strategies specific to eating behavior.This study is not without limitations, and these limitations may have influenced the findings. For example, the Tri-Trac is an external device that was worn continuously by the participants during the observation week. The presence of this device may have been a constant reminder to the subjects that they were being observed, and this may have prompted some subjects to be as accurate as possible when recording their exercise sessions or to increase their activity level. This reflects a significant difference from studies that have used doubly labeled water, a measuring technique that may be less noticeable to the subjects during the 1-week observation period. However, self-reported exercise did not increase on weeks that the Tri-Trac was worn compared to the other weeks during the weight loss program, suggesting that wearing the Tri-Trac did not influence the overall level of activity.This study primarily examined the accuracy of self-reported exercise, which is only one component of the energy balance equation. Weight regulation can also be influenced by dietary intake, and Litchman et al.(6) have shown that overweight adults under-report the number of calories that they consume. The self-reported dietary intake data presented in Table 5 suggests that there was no difference in the compliance to the diet between individuals who either over- or under-reported their exercise. However, it is not known whether the self-reported dietary information is accurate and whether inaccuracies in self-reported intake predicted weight loss as well or better than inaccuracies in self-reported exercise. Doubly labeled water would be required to determine the accuracy with which subjects reported their dietary intake.This study also has a number of strengths. For example, subjects in this study were taught how to self monitor their activity, and they had at least 4 wk to develop these skills before wearing the Tri-Trac. In contrast, it appears that other studies investigating the accuracy of self-reported exercise have tended to provide minimal training to individuals on how to self-monitor accurately (5) or have used techniques in which participants report their activity retrospectively. Therefore, it is possible that this training could markedly improve one's ability to report accurately the amount of exercise. Thus, it may be beneficial for future studies to examine the amount of training that is necessary for individuals to accurately self-monitor their exercise behaviors in a free living environment.In summary, it appears that approximately 40-60% of overweight women in a weight loss program over-report the amount of exercise they perform and these women lose less weight than those who under-report or accurately report their exercise (see Tables 3 and 4). Thus, the periodic use of an objective measure of exercise adherence (e.g., Tri-Trac) may allow these individuals who over-report their exercise to be identified so that appropriate strategies can be implemented to assist these individuals in developing their exercise behavior and improving their weight loss. Further, future studies should continue to examine predictors of compliance to behaviors related to weight loss, primarily exercise and diet, and to develop strategies that will improve compliance to these behaviors.This study was supported by the Obesity/Nutrition Research Center at the University of Pittsburgh which is funded by the National Institutes of Health(DK46204). The authors would like to thank Wei Lang for providing consultation on the statistical analyses for this project.Address for correspondence: John M. Jakicic, Ph.D., University of Pittsburgh School of Medicine, Western Psychiatric Institute and Clinic, 3811 O'Hara Street, Pittsburgh, PA 15213. E-mail:jakicicjm@msx.upmc.edu.REFERENCES1. Bouten, C. V., K. R. Westerterp, M. Verduin, and J. D. Janssen. Assessment of energy expenditure for physical activity using a triaxial accelerometer. Med. Sci. Sports Exerc. 26:1516-1523, 1994. [CrossRef] [Full Text] [Medline Link] [Context Link]2. Hagan, R. D., S. J. Upton, and J. Whittam. The effects of aerobic conditioning and/or caloric restriction in overweight men and women.Med. Sci. Sports Exerc. 18:87-94, 1986. [CrossRef] [Full Text] [Medline Link] [Context Link]3. Hill, J. O., D. G. Schlundt, T. Sbrocco, T. Sharp, J. Pope-Cordle, B. Stetson, M. Kaler, and C. Heim. Evaluation of an alternating-calorie diet with and without exercise in the treatment of obesity. Am. J. Clin. Nutr. 50:248-254, 1989. [Context Link]4. Jakicic, J. M., R. R. Wing, B. A. Butler, and R. J. Robertson. Prescribing exercise in multiple short bouts versus one continuous bout: effects on adherence, cardiorespiratory fitness, and weight loss in overweight women. Int. J. Obes. 19:893-901, 1995. [Medline Link] [Context Link]5. King, A. C., W. L. Haskell, C. B. Taylor, H. C. Kraemer, and R. F. DeBusk. Group- vs home-based exercise training in healthy older men and women. J. A. M. A. 266:1535-1542, 1991. [Context Link]6. Lichtman, S. W., K. Pisarska, E. R. Berman, et al. Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. N. Engl. J. Med. 327:1893-1898, 1992. [Context Link]7. Matthews, C. E. and P. S. Freedson. Field trial of a three-dimensional activity monitor: comparison with self report. Med. Sci. Sports Exerc. 27:1071-1078, 1995. [CrossRef] [Full Text] [Medline Link] [Context Link]8. Metropolitan Life Insurance Company. Nutrition and Athletic Performance: Values for Ideal Weight. New York: Metropolitan Life Insurance, 1983. [Context Link]9. Pronk, N. P. and R. R. Wing. Physical activity and long-term maintenance of weight loss. Obes. Res. 2:587-599, 1994. [Context Link]10. Welk G. J. and C. B. Corbin. The validity of the Tritrac-R3D activity monitor for the assessment of physical activity in children. Res. Q. Exerc. Sport. 66:202-209, 1995. [CrossRef] [Medline Link] [Context Link] WEIGHT LOSS; OBESITY; PHYSICAL ACTIVITY; EXERCISE ADHERENCEovid.com:/bib/ovftdb/00005768-199804000-0002400005768_1994_26_1516_bouten_accelerometer_|00005768-199804000-00024#xpointer(id(R1-24))|11065213||ovftdb|00005768-199412000-00016SL00005768199426151611065213P46[CrossRef]10.1249%2F00005768-199412000-00016ovid.com:/bib/ovftdb/00005768-199804000-0002400005768_1994_26_1516_bouten_accelerometer_|00005768-199804000-00024#xpointer(id(R1-24))|11065404||ovftdb|00005768-199412000-00016SL00005768199426151611065404P46[Full Text]00005768-199412000-00016ovid.com:/bib/ovftdb/00005768-199804000-0002400005768_1994_26_1516_bouten_accelerometer_|00005768-199804000-00024#xpointer(id(R1-24))|11065405||ovftdb|00005768-199412000-00016SL00005768199426151611065405P46[Medline Link]7869887ovid.com:/bib/ovftdb/00005768-199804000-0002400005768_1986_18_87_hagan_conditioning_|00005768-199804000-00024#xpointer(id(R2-24))|11065213||ovftdb|00005768-198602000-00015SL000057681986188711065213P47[CrossRef]10.1249%2F00005768-198602000-00015ovid.com:/bib/ovftdb/00005768-199804000-0002400005768_1986_18_87_hagan_conditioning_|00005768-199804000-00024#xpointer(id(R2-24))|11065404||ovftdb|00005768-198602000-00015SL000057681986188711065404P47[Full Text]00005768-198602000-00015ovid.com:/bib/ovftdb/00005768-199804000-0002400005768_1986_18_87_hagan_conditioning_|00005768-199804000-00024#xpointer(id(R2-24))|11065405||ovftdb|00005768-198602000-00015SL000057681986188711065405P47[Medline Link]3457234ovid.com:/bib/ovftdb/00005768-199804000-0002400004348_1995_19_893_jakicic_cardiorespiratory_|00005768-199804000-00024#xpointer(id(R4-24))|11065405||ovftdb|SL0000434819951989311065405P49[Medline Link]8963358ovid.com:/bib/ovftdb/00005768-199804000-0002400005768_1995_27_1071_matthews_dimensional_|00005768-199804000-00024#xpointer(id(R7-24))|11065213||ovftdb|00005768-199507000-00017SL00005768199527107111065213P52[CrossRef]10.1249%2F00005768-199507000-00017ovid.com:/bib/ovftdb/00005768-199804000-0002400005768_1995_27_1071_matthews_dimensional_|00005768-199804000-00024#xpointer(id(R7-24))|11065404||ovftdb|00005768-199507000-00017SL00005768199527107111065404P52[Full Text]00005768-199507000-00017ovid.com:/bib/ovftdb/00005768-199804000-0002400005768_1995_27_1071_matthews_dimensional_|00005768-199804000-00024#xpointer(id(R7-24))|11065405||ovftdb|00005768-199507000-00017SL00005768199527107111065405P52[Medline Link]7564975ovid.com:/bib/ovftdb/00005768-199804000-0002400006966_1995_66_202_welk_assessment_|00005768-199804000-00024#xpointer(id(R10-24))|11065213||ovftdb|SL0000696619956620211065213P55[CrossRef]10.1080%2F02701367.1995.10608834ovid.com:/bib/ovftdb/00005768-199804000-0002400006966_1995_66_202_welk_assessment_|00005768-199804000-00024#xpointer(id(R10-24))|11065405||ovftdb|SL0000696619956620211065405P55[Medline Link]7481081Accuracy of self-reported exercise and the relationship with weight loss in overweight womenJAKICIC, JOHN M.; POLLEY, BETSY A.; WING, RENA R.Special Communications: Methods430InternalMedicine & Science in Sports & Exercise1999315747-754MAY 1999The accuracy of the TriTrac-R3D accelerometer to estimate energy expenditureJAKICIC, JM; WINTERS, C; LAGALLY, K; HO, J; ROBERTSON, RJ; WING, RRhttp://journals.lww.com/acsm-msse/Fulltext/1999/05000/The_accuracy_of_the_TriTrac_R3D_accelerometer_to.20.aspxhttp://pdfs.journals.lww.com/acsm-msse/1999/05000/The_accuracy_of_the_TriTrac_R3D_accelerometer_to.00020.pdfInternalMedicine & Science in Sports & Exercise2000325976-984MAY 2000Comparing physical activity assessment methods in the Seasonal Variation of Blood Cholesterol StudyMATTHEWS, CE; FREEDSON, PS; HEBERT, JR; STANEK, EJ; MERRIAM, PA; OCKENE, IShttp://journals.lww.com/acsm-msse/Fulltext/2000/05000/Comparing_physical_activity_assessment_methods_in.15.aspx208http://pdfs.journals.lww.com/acsm-msse/2000/05000/Comparing_physical_activity_assessment_methods_in.00015.pdfInternalMedicine & Science in Sports & Exercise2004365897-904MAY 2004Evaluation of the SenseWear Pro Armband™ to Assess Energy Expenditure during ExerciseJAKICIC, JM; MARCUS, M; GALLAGHER, KI; RANDALL, C; THOMAS, E; GOSS, FL; ROBERTSON, RJhttp://journals.lww.com/acsm-msse/Fulltext/2004/05000/Evaluation_of_the_SenseWear_Pro_Armband__to_Assess.24.aspx241http://pdfs.journals.lww.com/acsm-msse/2004/05000/Evaluation_of_the_SenseWear_Pro_Armband__to_Assess.00024.pdfInternalMedicine & Science in Sports & Exercise20003271320-1326JUL 2000Comparisons of four methods of estimating physical activity in adult womenLEENDERS, NY; SHERMAN, WM; NAGARAJA, HNhttp://journals.lww.com/acsm-msse/Fulltext/2000/07000/Comparisons_of_four_methods_of_estimating_physical.21.aspx95http://pdfs.journals.lww.com/acsm-msse/2000/07000/Comparisons_of_four_methods_of_estimating_physical.00021.pdfInternalMedicine & Science in Sports & Exercise20033581374-1380AUG 2003The Validity of the Stanford Seven-Day Physical Activity Recall in Young AdultsWASHBURN, RA; JACOBSEN, DJ; SONKO, BJ; HILL, JO; DONNELLY, JEhttp://journals.lww.com/acsm-msse/Fulltext/2003/08000/The_Validity_of_the_Stanford_Seven_Day_Physical.19.aspx173http://pdfs.journals.lww.com/acsm-msse/2003/08000/The_Validity_of_the_Stanford_Seven_Day_Physical.00019.pdfInternalMedicine & Science in Sports & Exercise20033591605-1611SEP 2003Estimating Physical Activity Using the CSA Accelerometer and a Physical Activity LogSCHMIDT, MD; FREEDSON, PS; CHASAN-TABER, Lhttp://journals.lww.com/acsm-msse/Fulltext/2003/09000/Estimating_Physical_Activity_Using_the_CSA.23.aspx162http://pdfs.journals.lww.com/acsm-msse/2003/09000/Estimating_Physical_Activity_Using_the_CSA.00023.pdf