Ann M. Swartz

Compendium of physical activities: an update of activity codes and MET intensities.

We provide an updated version of the Compendium of Physical Activities, a coding scheme that classifies specific physical activity (PA) by rate of energy expenditure. It was developed to enhance the comparability of results across studies using self-reports of PA. The Compendium coding scheme links a five-digit code that describes physical activities by major headings (e.g., occupation, transportation, etc.) and specific activities within each major heading with its intensity, defined as the ratio of work metabolic rate to a standard resting metabolic rate (MET). Energy expenditure in MET-minutes, MET-hours, kcal, or kcal per kilogram body weight can be estimated for specific activities by type or MET intensity. Additions to the Compendium were obtained from studies describing daily PA patterns of adults and studies measuring the energy cost of specific physical activities in field settings. The updated version includes two new major headings of volunteer and religious activities, extends the number of specific activities from 477 to 605, and provides updated MET intensity levels for selected activities.

Validity of four motion sensors in measuring moderate intensity physical activity

PURPOSEnThis study tested the validity of four motion sensors for measuring energy expenditure (EE) during moderate intensity physical activities in field and laboratory settings. We also evaluated the accuracy of the EE values for selected moderate activities listed in the 1993 Compendium of Physical Activities.nnnMETHODSnA total of 81 participants (age 19-74 yr) completed selected tasks from six general categories: yardwork, housework, occupation, family care, conditioning, and recreation. Twelve individuals performed each of the 28 activities examined. During each activity, EE was measured using a portable metabolic measurement system. Participants also wore three accelerometers (Computer Science and Applications [CSA], Inc. model 7164; Caltrac; and Kenz Select 2) and the Yamax SW-701 electronic pedometer. For the CSA device, three previously developed regression equations were used to convert accelerometer scores to EE.nnnRESULTSnThe mean error scores (indirect calorimetry minus device) across all activities were: CSA1, 0.97 MET; CSA2, 0.47 MET, CSA3, 0.05 MET; Caltrac, 0.83 MET; Kenz, 0.96 MET; and Yamax, 1.12 MET. The correlation coefficients between indirect calorimetry and motion sensors ranged from r = 0.33 to r = 0.62. The energy cost for power mowing and sweeping/mopping was higher than that listed in the 1993 Compendium (P < 0.05), and the cost for several household and recreational activities was lower (P < 0.05).nnnCONCLUSIONnMotion sensors tended to overpredict EE during walking. However, they underpredicted the energy cost of many other activities because of an inability to detect arm movements and external work. These findings illustrate some of the limitations of using motion sensors to predict EE in field settings.

Estimation of energy expenditure using CSA accelerometers at hip and wrist sites

PURPOSEnThis study was designed to establish prediction models that relate hip and wrist accelerometer data to energy expenditure (EE) in field and laboratory settings. We also sought to determine whether the addition of a wrist accelerometer would significantly improve the prediction of EE (METs), compared with a model that used a hip accelerometer alone.nnnMETHODSnSeventy participants completed one to six activities within the categories of yardwork, housework, family care, occupation, recreation, and conditioning, for a total of 5 to 12 participants tested per activity. EE was measured using the Cosmed K4b2 portable metabolic system. Simultaneously, two Computer Science and Applications, Inc. (CSA) accelerometers (model 7164), one worn on the wrist and one worn on the hip, recorded body movement. Correlations between EE measured by the Cosmed and the counts recorded by the CSA accelerometers were calculated, and regression equations were developed to predict EE from the CSA data.nnnRESULTSnThe wrist, hip, and combined hip and wrist regression equations accounted for 3.3%, 31.7%, and 34.3% of the variation in EE, respectively. The addition of the wrist accelerometer data to the hip accelerometer data to form a bivariate regression equation, although statistically significant (P = 0.002), resulted in only a minor improvement in prediction of EE. Cut points for 3 METs (574 hip counts), 6 METs (4945 hip counts), and 9 METs (9317 hip counts) were also established.nnnCONCLUSIONnThe small amount of additional accuracy gained from the wrist accelerometer is offset by the extra time required to analyze the data and the cost of the accelerometer.

Guide to the Assessment of Physical Activity: Clinical and Research Applications A Scientific Statement From the American Heart Association

The deleterious health consequences of physical inactivity are vast, and they are of paramount clinical and research importance. Risk identification, benchmarks, efficacy, and evaluation of physical activity behavior change initiatives for clinicians and researchers all require a clear understanding of how to assess physical activity. In the present report, we have provided a clear rationale for the importance of assessing physical activity levels, and we have documented key concepts in understanding the different dimensions, domains, and terminology associated with physical activity measurement. The assessment methods presented allow for a greater understanding of the vast number of options available to clinicians and researchers when trying to assess physical activity levels in their patients or participants. The primary outcome desired is the main determining factor in the choice of physical activity assessment method. In combination with issues of feasibility/practicality, the availability of resources, and administration considerations, the desired outcome guides the choice of an appropriate assessment tool. The decision matrix, along with the accompanying tables, provides a mechanism for this selection that takes all of these factors into account. Clearly, the assessment method adopted and implemented will vary depending on circumstances, because there is no single best instrument appropriate for every situation. In summary, physical activity assessment should be considered a vital health measure that is tracked regularly over time. All other major modifiable cardiovascular risk factors (diabetes mellitus, hypertension, hypercholesterolemia, obesity, and smoking) are assessed routinely. Physical activity status should also be assessed regularly. Multiple physical activity assessment methods provide reasonably accurate outcome measures, with choices dependent on setting-specific resources and constraints. The present scientific statement provides a guide to allow professionals to make a goal-specific selection of a meaningful physical activity assessment method.

Comparison of three methods for measuring the time spent in physical activity.

PURPOSEnThree methods for measuring time spent in daily physical activity (PA) were compared during a 21-d period among 83 adults (38 men and 45 women).nnnMETHODSnEach day, participants wore a Computer Science and Applications, Inc. (CSA) monitor and completed a 1-page, 48-item PA log that reflected time spent in household, occupational, transportation, sport, conditioning, and leisure activities. Once a week, participants also completed a telephone survey to identify the number of minutes spent each week in nonoccupational walking and in moderate intensity and hard/very hard-intensity PA. Data were analyzed using descriptive statistics and Spearman rank-order correlations. Three equations developed to compute CSA cut points for moderate and hard/very hard PA were also compared with the PA logs and PA survey.nnnRESULTSnThere was modest to good agreement for the time spent in different PA intensity categories between the three CSA cut point methods (r = 0.43-0.94, P < 0.001). Correlations between the CSA and PA logs ranged from r = 0.22 to r = 0.36, depending on the comparisons. Correlations between the survey items and PA logs were r = 0.26-0.54 (P < 0.01) for moderate and walking activities and r < 0.09 (P > 0.05) for hard/very hard activities. Correlations between the survey items and the CSA min per day varied according to the method used to compute the CSA intensity cut points.nnnCONCLUSIONSnThe results were consistent with findings from other PA validation studies that show motion sensors, PA logs, and surveys reflect PA; however, these methods do not always provide similar estimates of the time spent in resting/light, moderate, or hard/very hard PA.

A meta-analysis of pedometer-based walking interventions and weight loss

PURPOSE Cross-sectional studies show that individuals who walk more tend to be thinner than those who walk less. This does not mean, however, that the association between higher step counts and lower weight is causal or that encouraging sedentary individuals to increase step counts helps them lose weight. METHODS In this meta-analysis, we searched 6 electronic databases and contacted pedometer experts to identify pedometer-based walking studies without a dietary intervention that reported weight change as an outcome. We included randomized controlled trials and prospective cohort studies published after January 1, 1995, in either English or Japanese, with 5 or more adult participants and at least 1 cohort enrolled in a pedometer-based walking intervention lasting at least 4 weeks. RESULTS Nine studies met the study inclusion criteria. Cohort sample size ranged from 15 to 106, for a total of 307 participants, 73% of whom were women and 27% of whom were men. The duration of the intervention ranged from 4 weeks to 1 year, with a median duration of 16 weeks. The pooled estimate of mean weight change from baseline using a fixed-effects model and combining data from all 9 cohorts was −1.27 kg (95% confidence interval, −1.85 to −0.70 kg). Longer intervention duration was associated with greater weight change. On average, participants lost 0.05 kg per week during the interventions. CONCLUSION Pedometer-based walking programs result in a modest amount of weight loss. Longer programs lead to more weight loss than shorter programs.

Relationship of leisure-time physical activity and occupational activity to the prevalence of obesity

OBJECTIVE: To assess the interaction between leisure-time physical activity (LTPA) and occupational activity (OA) on the prevalence of obesity.DESIGN: Secondary data analysis of a population based cross-sectional US national sample (NHANES III).SUBJECTS: A total of 4889 disease-free, currently employed adults over age 20u2005y.MEASUREMENTS: Subjects body mass index (BMI) was categorized as (1) obese (BMI≥30u2005kg/m2), or (2) non-obese (BMI<30u2005kg/m2). LTPA was divided into four categories: (1) no LTPA; (2) irregular LTPA; (3) regular moderate intensity LTPA; and (4) regular vigorous intensity LTPA. OA was grouped as (1) high OA and (2) low OA. Age, gender, race–ethnicity, smoking status, urbanization classification, alcohol consumption and income were statistically controlled.RESULTS: In all, 16.8% (s.e. 0.7) of the total subject population were obese (15.1% (s.e. 1.1) of men and 19.1% (s.e. 1.1) of women). Logistic regression revealed that compared to those who engage in no LTPA and have low levels of OA, the likelihood of being obese is 42% (95% CI 0.35, 0.96) lower for those who engage in no LTPA and have high OA, 48% (95% CI 0.32, 0.83) lower for those who have irregular LTPA and have high levels of OA, and about 50% lower for all those who have regular LTPA through moderate or vigorous activity levels regardless of OA level.CONCLUSION: When considering disease free adults above 20u2005y of age employed in high and low activity occupations, a high level of occupational activity is associated with a decreased likelihood of being obese.

Evaluation of heart rate as a method for assessing moderate intensity physical activity

UNLABELLEDnTo further develop our understanding of the relationship between habitual physical activity and health, research studies require a method of assessment that is objective, accurate, and noninvasive. Heart rate (HR) monitoring represents a promising tool for measurement because it is a physiological parameter that correlates well with energy expenditure (EE). However, one of the limitations of HR monitoring is that training state and individual HR characteristics can affect the HR-VO2 relationship.nnnPURPOSEnThe primary purpose of this study was to examine the relationship between HR (beats x min(-1)) and VO2 (mL x kg(-1 x -1) min(-1)) during field- and laboratory-based moderate-intensity activities. In addition, we examined the validity of estimating EE from HR after adjusting for age and fitness. This was done by expressing the data as a percent of heart rate reserve (%HRR) and percent of VO2 reserve (%VO2R).nnnMETHODSnSixty-one adults (18-74 yr) performed physical tasks in both a laboratory and field setting. HR and VO2 were measured continuously during the 15-min tasks. Mean values over min 5-15 were used to perform linear regression analysis on HR versus VO2. HR data were then used to predict EE (METs), using age-predicted HRmax and estimated VO2max.nnnRESULTSnThe correlation between HR and VO2 was r = 0.68, with HR accounting for 47% of the variability in VO2. After adjusting for age and fitness level, HR was an accurate predictor of EE (r = 0.87, SEE = 0.76 METs).nnnCONCLUSIONnThis method of analyzing HR data could allow researchers to more accurately quantify physical activity in free-living individuals.

Increasing daily walking improves glucose tolerance in overweight women

BACKGROUNDnPhysical activity (PA) has been shown to benefit glucose tolerance. Walking is a convenient low-impact mode of PA and is reported to be the most commonly performed activity for those with diabetes. The purpose of this study was to determine whether a recommendation to accumulate 10,000 steps/day for 8 weeks was effective at improving glucose tolerance in overweight, inactive women.nnnMETHODSnEighteen women (53.3 +/- 7.0 years old, 35.0 +/- 5.1 kg/m(2)) with a family history of type 2 diabetes completed a 4-week control period followed by an 8-week walking program with no changes in diet. The walking program provided a goal of accumulating at least 10,000 steps/day, monitored by a pedometer.nnnRESULTSnDuring the control period, participants walked 4972 steps/day. During the intervention period, the participants increased their accumulated steps/day by 85% to 9213, which resulted in beneficial changes in 2-h postload glucose levels (P < 0.001), AUC(glucose) (P = 0.025), systolic blood pressure (P < 0.001), and diastolic blood pressure (P = 0.002). There were no changes in body mass, body fat percentage, and waist circumference during the walking intervention.nnnCONCLUSIONSnThe 10,000 steps/day recommendation resulted in improved glucose tolerance and a reduction in systolic and diastolic blood pressure in overweight women at risk for type 2 diabetes. This demonstrates that activity can be accumulated throughout the day and does not have to result in weight loss to benefit this population.

Simultaneous heart rate-motion sensor technique to estimate energy expenditure

PURPOSEnHeart rate (HR) and motion sensors represent promising tools for physical activity (PA) assessment, as each provides an estimate of energy expenditure (EE). Although each has inherent limitations, the simultaneous use of HR and motion sensors may increase the accuracy of EE estimates. The primary purpose of this study was to establish the accuracy of predicting EE from the simultaneous HR-motion sensor technique. In addition, the accuracy of EE estimated by the simultaneous HR-motion sensor technique was compared to that of HR and motion sensors used independently.nnnMETHODSnThirty participants (16 men: age, 33.1 +/- 12.2 yr; BMI, 26.1 +/- 0.7 kg.m(-2); and 14 women: age, 31.9 +/- 13.1 yr; BMI, 27.2 +/- 1.1 kg.m(-2) (mean +/- SD)) performed arm and leg work in the laboratory for the purpose of developing individualized HR-VO2 regression equations. Participants then performed physical tasks in a field setting for 15 min each. CSA accelerometers placed on the arm and leg were to discriminate between upper and lower body movement, and HR was then used to predict EE (METs) from the corresponding arm or leg laboratory regression equation. A hip-mounted CSA accelerometer and Yamax pedometer were also used to predict EE. Predicted values (METs) were compared to measured values (METs), obtained via a portable metabolic measurement system (Cosmed K4b(2)).nnnRESULTSnThe Yamax pedometer and the CSA accelerometer on the hip significantly underestimated the energy cost of selected physical activities, whereas HR alone significantly overestimated the energy cost of selected physical activities. The simultaneous HR-motion sensor technique showed the strongest relationship with VO(2) (R(2) = 0.81) and did not significantly over- or underpredict the energy cost (P = 0.341).nnnCONCLUSIONnThe simultaneous HR-motion sensor technique is a good predictor of EE during selected lifestyle activities, and allows researchers to more accurately quantify free-living PA.