Moderate levels of regular physical activity provide health benefits associated with reduced risks of disease and premature death (1). Based on national physical activity surveillance system data, most women, especially those from ethnic minorities, those who are ≥ 65 yr old, and those with low educational attainment are not regularly active at levels that confer health benefits. According to the 1996 Surgeon General’s Report on Physical Activity, nearly one third of women ages 45 yr and older do not participate in leisure-time physical activities (22) Inactivity rates increase with age, ranging from 25% to 51% between ages 18 to 75 yr and older, and are higher among women of color (from 39 to 43%) than white women (28%). Accordingly, the report concludes that “Physical inactivity is more prevalent among women than men, among Blacks and Hispanics than among whites, among older than younger adults, and among the less affluent than the more affluent” (22).
With the exception of walking and gardening activities, types of physical activities measured on surveillance system surveys focus on sports and conditioning activities (2). Studies using physical activity records and timed diaries suggest women spend less time in organized sports and heavy-intensity exercises and more time in housework, caregiving activities, and occupation activities (3,15). Omitting household, family care, and occupational activities from physical activity surveys may underestimate activity patterns among women who do not pursue recreational and conditioning activities during their leisure time.
In 1997, Kriska and Caspersen (13) published a monograph of physical activity surveys used in public health research. Among the 31 surveys listed, about half had items about occupational activities, six had items about housework, and only two surveys had items about caring for others. None of the surveys asked respondents about the time they spent in family care activities. In 1998, Sternfeld et al. (19) published a study of physical activity patterns among 2636 ethnically diverse women enrolled in the Northern California Kaiser Permanente Medical Care Program. The investigators used the Kaiser Physical Activity Survey (KPAS) to measure correlates of housework/caregiving, sports/exercise, active living habits, and occupation activities. That study found that the demographic and psychosocial factors associated with a high level of physical activity varied by domain. For instance, women who had the highest level of participation in sports and exercise and active-living behaviors were more likely to be young, white, college-educated, and without young children at home. They also tended to have high self-efficacy and social support for exercise, and did not suffer from lack of motivation or have perceived barriers to exercise. In contrast, women with a high level of household/caregiving activity were more likely to be older, married, of Hispanic ethnicity, to have young children at home and to perceive time constraints as a barrier to exercise, whereas women with a high level of occupational activity tended to be less educated and to be current cigarette smokers.
In this study, we present the results from a study designed to evaluate the 1-month test-retest reliability of the KPAS and to compare the survey items and summary scores with direct and indirect measures of physical activity. We compare the scores for housework/caregiving, work, active living habits, and sports/exercise related activities listed in the KPAS with both direct and indirect criterion measures of physical activity in a sample of 50 women with a broad range of physical activity habits.
MATERIALS AND METHODS
Inclusion criteria were established to enroll women in the KPAS evaluation study who were similar in age with those in the Northern California Kaiser Permanente Medical Care Program cohort and who were able to complete the study activities. Eligibility requirements included: (a) the absence of any conditions that would prevent participants from completing a maximal treadmill test or performing the activities of daily living (5), (b) being between ages 18 and 60 yr, (c) the ability to keep detailed records of physical activity, and (d) the willingness to comply with the universities institutional review board for studies involving human subjects. Presence of health conditions that would exclude participants from the study was identified by a health history survey completed by the participants and a cardiovascular and respiratory physical exam performed by the study investigators. Participants were recruited by advertisements placed in the universities academic and hospital facilities. Fifty-three women, ages 18–60 yr, were initially enrolled in the study and 50 women completed the study protocol. Of the three nonfinishers, all dropped out after the first visit because of time constraints. Data from 50 women are included in this report. Among the participants finishing the study, most were white (94%), college educated (78%), and employed full time (70%). About half the sample were married (48%), with the rest being never married (34%) or divorced, separated, or widowed (18%). There were no apparent differences in age, education, employment, physical fitness, and physical activity measures between those who did and did not complete the study.
Participants of the KPAS evaluation study completed four clinic visits, spaced 1–3 wk apart. Before enrolling in the study, volunteers were screened for inclusion criteria by telephone or in person. Those that met the criteria were scheduled for the first clinic visit, approximately 1 wk later. During the first visit, participants read and signed an informed consent form approved by the University of North Carolina Institutional Review Board for Research Involving Human Subjects, performed a maximal treadmill graded exercise test to determine their cardiorespiratory fitness, completed the KPAS, and received instructions for completing the physical activity record (PA record) and wearing the Caltrac accelerometer. During the next 7 d, participants kept detailed records of all physical activities in a book developed for this study and simultaneously wore a Caltrac accelerometer to obtain direct measure of their daily physical activity. The second clinic visit was scheduled 1 wk after the first visit. During the second visit, study staff reviewed the completed PA record and Caltrac accelerometer data with the participants and clarified any unclear or missing data. Participants were also weighed underwater to determine their percent body fat. The third and fourth clinic visits were scheduled 1 month after clinic visits 1 and 2. During the third visit, participants completed the KPAS questionnaire for a second time and were instructed once again to record their physical activity in a record book and wear a Caltrac accelerometer during the following week. One week later, participants returned to the clinic for their fourth and last visit. Study staff reviewed the completed PA record and Caltrac accelerometer data with the participants and clarified any unclear or missing data.
Kaiser Physical Activity Survey.
The KPAS is a self-administered eight-page instrument designed to obtain information about women’s physical activity habits. The survey contains 75 items and takes about 20 min to complete. The survey has seven sections: housework/caregiving, occupation, active living habits, sports/exercise activities, personal feelings about exercise, contemplation about exercise, and personal characteristics. The first four sections are used to classify physical activity status. With the exception of the caregiving section, summary indexes are computed from five-level categorical responses to questions about participation in various activities. Responses range from 1 for “never” to 5 for “always.” For the sports/exercise section, respondents also are asked to identify the frequency and duration for the three most frequent sports/exercise activities performed in the past yr. For the caregiving section, four-level categorical responses, ranging from 1 for “none” to 4 for “20 h or more per week” reflect the time per week spent in caregiving activities. A detailed description of the KPAS survey and scoring procedures is given by Sternfeld et al. (19). A copy of the survey is presented in Appendix 1.
KPAS activity indexes were computed for the housework, caregiving, occupation, active living habits, and sports/exercise survey items using the scoring procedures outlined by Sternfeld et al. (19) and Baecke et al. (6). A three-item summary index proposed by Baecke et al. (6) was computed by summing the sports/exercise, activity living habits and occupation indexes. This differed from the three-item summary index computed by Sternfeld et al. (19), who combined the sports/exercise, activity living habits, and housework/caregiving indexes since some women were unemployed and did not have an occupation index score. A four-item summary index was computed by adding the housework/caregiving index to the KPAS three-item summary index.
Physical activity criterion measures.
The KPAS survey items and summary scores were compared with both direct (physical activity records and motion detectors) and indirect (cardiorespiratory fitness and percent body fat) criterion measures associated with habitual physical activity. The direct measures of physical activity records and motion detectors were selected because they provide a qualitative measure of the type, duration, and perceived intensity of physical activities performed (physical activity records) and a quantitative measure of physical activity in activity counts or estimated energy expenditure (motion detectors). In this study, the physical activity records and the motion detectors were worn concurrently, giving a simultaneous measure of the quality and quantity of physical activity. The indirect measures of cardiorespiratory fitness and percent body fat were selected as criterion measures of physical activity because they provide physiological measures of habitual physical activity and are related with all-cause mortality and some causes of morbidity (22).
Physical activity records (PA records).
For two 7-d periods, scheduled 1 month apart, participants recorded all physical activity in a PA record designed for this study. Participants were asked to be as precise and conscientious as possible and to make PA record entries of all activities during this period as conditions warranted and/or permitted. Participants recorded the following information for each activity performed in the PA record: type and description; mode (reclining, sitting, standing, walking, or running); subjective estimate of the intensity (light, moderate, or vigorous) or the pace (slow, moderate, or brisk); and the duration in hours and minutes. Types of activities recorded included, but were not limited to, sports and recreational activities, household chores, shopping, cooking, personal care and hygiene activities, child care, activities of daily living, sleeping, transportation, lawn and garden activities, and occupational tasks. After PA records were completed, trained interviewers edited the records for clarity and accuracy with the participants. Interviewers then assigned each activity a five-digit code from the Compendium of Physical Activity that reflects the type and MET level intensity of activity performed (4). A MET is the ratio of the activity metabolic rate to the resting metabolic rate. Physical activity energy expenditure was computed by multiplying the MET intensities for each activity by min per event (MET-min) and summing the MET-min accumulated each day (MET-min·d−1). Data obtained from the two 7-d PA records were averaged to present a MET-min·d−1 score averaged across 2 wk.
Summary scores were computed from the PA records to compare with the KPAS indexes for sports/exercise, housework/caregiving, active living habits, and occupation activities. We computed summary scores for the PA records using similar types of activities as listed in the KPAS. When we were unable to link activities for index items in the KPAS with like activities in the PA record, direct comparisons were not made.
The PA record sports/exercise score was computed by adding the MET-min·d−1 scores for bicycling, conditioning exercises, dancing, fishing and hunting, jogging, sports, recreational walking, water, and winter activities. The intensity of the physical recreation activities ranged from 2.5 to 18 METs. The PA record housework/caregiving score was computed by adding the MET-min·d−1 scores for child care, animal care, care for dependent adults, cooking, house cleaning, laundry, grocery shopping, other shopping, and other miscellaneous household activities. Intensities for the housework activities ranged from 1.2 to 8.0 METs. Intensities for the caregiving activities ranged from 1.5 to 5.0 METs. The PA record active living habits score was computed by adding the MET-min·d−1 scores for walking and/or bicycling for leisure and/or transportation and subtracting the MET-min·d−1 score for watching television. Negative scores reflected more time and energy spent in watching television than in walking and bicycling. Intensities for the active living habits activities ranged from 1.0 to 4.0 METs. The PA record occupation summary score was computed by adding the MET-min·d−1 scores for all occupational activities listed in the PA records. A three-item summary score was computed by adding the PA record MET-min·d−1 scores for occupation, sports/exercise, and active living habits. A four-item summary score was computed by adding the score for the housework/caregiving MET-min·d−1 to the three-item summary score.
The Caltrac Accelerometer (Muscle Dynamics, Torrance, CA) is a light-weight electronic device, equipped with a microcomputer, which is used to assess daily body movement by recording vertical accelerations of the body. Acceleration counts are converted internally to an estimate of overall energy expenditure based on estimated resting energy expenditure derived from the subject’s height, weight, sex, and age (10). The Caltrac was worn on the participant’s hip for the same 7 d that the PA records were maintained, excluding sleep, showering, and water sports. At the end of each day, participants wrote the kcal value presented on the digital display of the Caltrac accelerometer in the PA record. In the event that the Caltrac malfunctioned (N = 4 times), participants were given a new accelerometer to resume activity monitoring the next day. Also, in cases when subjects did not wear the accelerometer for one or more days (N = 2), average daily Caltrac scores were computed as the mean of the days per week each subject wore the accelerometer. Caltrac 24-h scores are presented as Caltrac kcal·d−1, which factors in individual estimated resting energy expenditure, and as Caltrac MET-min·d−1, which is independent of resting metabolism. Caltrac MET-min·d−1 scores were computed by dividing each participant’s Caltrac kcal score by their estimated resting metabolic rate and multiplying by 1440 min·d−1 (1). The equation of Mifflin et al. (11) was used to compute the metabolic rate.
Cardiorespiratory fitness was assessed by direct measurement of oxygen uptake (O2) during a treadmill graded exercise test. The protocol used begins at 3 mph and 0% grade and progressively increases in grade or speed at an energy cost of about 1 MET per 2-min stage. Blood pressure and heart rate recordings were obtained immediately before the treadmill test, during the last 30 s of each stage, and during recovery by a trained technician. Heart rate and rhythm were monitored throughout the exercise test using a 12-lead ECG monitor. During each exercise stage, volume and concentrations of oxygen and carbon dioxide in the expired air were measured continuously using an automated Rayfield system (Rayfield Instruments, Waitsfield, VT). Exercise test endpoints were exhaustion, dyspnea, ECG abnormalities (ST-segment depression, VPB pairs or runs), signs or symptoms of exercise intolerance, or equipment technical problems (13). Because five women did not reach the classic criteria for O2max (completion of two successive stages yielding increases in oxygen uptake not greater than 2 mL·kg−1 ·min−1, a heart rate equal to or greater than 95% predicted maximal levels, and a respiratory exchange ratio greater than 1) (20), but exercised to voluntary exhaustion, we use the term O2 peak to reflect cardiorespiratory fitness.
Percent body fat was measured from body density determined by hydrostatic weighing and the Siri equation (18). Residual lung volume was measured by the closed circuit oxygen dilution method of Wilmore (23) and corrected for in the body fat determination. Height was measured in cm at the initial clinic visit using a wall-mounted tape measure. Body weight was measured in kilograms at each clinic visit using a calibrated laboratory scale.
Means ± standard deviations (SD) were computed for the study variables. Because the PA record and KPAS data were positively skewed, the median and 25th–75th quartile scores are presented for these variables. KPAS scores were compared using an age-adjusted Spearman rho correlation between the KPAS items and the direct and indirect physical activity measures. Data were age-adjusted because age is inversely associated with physical activity, cardiorespiratory fitness, and percent body fat. Survey items included the housework/caregiving, occupation, active living habits, and sports/exercise survey items and indexes. Direct and indirect measures of physical activity measures were the PA records in MET-min·d−1, Caltrac accelerometer in kcal·d−1 and MET-min·d−1, O2 peak in mL·kg−1min−1, and percent body fat. One-month test-retest reliability of the KPAS scores was determined using intraclass correlations. The intraclass correlation gives a measure of the relative homogeneity of the test-retest survey scores within subjects in relation to the total variation between subjects (8). A sample size of 50 is sufficient to detect significant correlations of r = 0.28 at an alpha of 0.05 and a beta of 0.85 and r = 0.35 at an alpha of 0.01 and a beta of 0.85. Analyses were completed using PC-SAS statistical software (Cary, NC).
Table 1 shows the descriptive data for the study variables. The mean ± SD for age of the study participants was 39.1 ± 12 yr. Mean ± SD values for cardiorespiratory fitness and percent body fat levels were 36.8 ± 10 mL·kg−1min−1 and 30.3 ± 10.2%, respectively. There was little difference between mean ± SD values of the Caltrac accelerometer scores expressed as kcal·d−1 (2154 ± 286) and MET-min·d−1 (2080 ± 260). The median and 25th–75th percentile scores are presented for the PA record and survey physical activity data. The median daily energy expenditure from the PA records was 2,464 MET-min·d−1. Occupation and housework/caregiving summary scores were higher (625 and 356 MET-min·d−1, respectively) than sports/exercise and active living habit summary scores (171 and 63.7 MET-min·d−1, respectively). Addition of the housework/caregiving activities to the three-item summary score increased the median daily energy expenditure from 800 MET-min·d−1 to 1260 MET-min·d−1. KPAS indexes were higher for sports/exercise and active living habits (4.1 and 3.0 units, respectively) as compared with occupation and housework/caregiving scores (2.3 and 2.2 units, respec- tively). Addition of the housework/caregiving activities to the KPAS three-item summary index increased the summary index from 9.0 to 11.1 units.
Table 2 shows the intraclass correlations for the 1-month test-retest reliability of the KPAS indexes. These correlations were generally high, ranging from ICC = 0.79 to 0.91 (P < 0.001).
Table 3 shows the age-adjusted Spearman correlations between the KPAS indexes and the physical activity comparison criteria. All KPAS indexes, except active living habits, were significantly related to corresponding PA record summary scores (r ≥ 0.35, P < 0.01). KPAS housework/caregiving and occupation indexes were related to the Caltrac kcal·d−1 scores (r ≥ 0.35, P < 0.05). KPAS sports/exercise and active living habits indexes were related to O2 peak (r ≥ 0.34, P < 0.01), percent body fat (r ≥ −0.41, P < 0.01), and the Caltrac MET-min·d−1 (r ≥ 0.30, P < 0.01). Other correlations between the KPAS indexes and physical activity comparison measures were low and not significant (r < 0.22, P > 0.05). Correlations between the KPAS summary indexes and all physical activity comparison measures, except the Caltrac kcal·d−1, were ≥ 0.35 (P < 0.01).
Table 4 shows the age-adjusted Spearman correlations between similar activities from the KPAS and PA records. Among the housework/caregiving items, correlations of ≥ 0.28 were obtained for child care, cooking, routine cleaning, gardening and yard work, and home repair activities (P < 0.05). None of the study participants reported elderly care activities. Among the occupation items, correlations of ≥ 0.36 were obtained for sitting, standing, walking, and sweating or feeling exertion at work (P < 0.01). Among the active living habits section, correlations of ≥ 0.33 were obtained for walking and bicycling ≥ 15 min per event and watching television (P < 0.05). The correlation for the corresponding indexes for sports/exercise activities performed regularly was r = 0.73.
These findings showed that, in our sample of women, the KPAS was a reliable instrument and was reasonably accurate in assessing certain types of activity. We observed correlations on the order of 0.50 or higher among KPAS indexes for housework, caregiving, and sports/exercise compared with similar items from two 7-d PA records. However, correlations were lower for the summary indexes for active living habits and occupation, ranging from 0.22 to 0.35.
Correlations between the KPAS indexes and the Caltrac accelerometer scores varied depending on how Caltrac scores were expressed. When the Caltrac scores were expressed as kcal·d−1 (largely reflecting components of the resting metabolic rate of body weight, height, age, and sex), correlations with the KPAS caregiving and occupation indexes were on the order of r ≥ 0.35. Conversely, when Caltrac scores were expressed as MET-min·d−1, correlations with the KPAS caregiving and occupation indexes were low and not significant. These differences in correlations, depending on how Caltrac data are expressed, may have reflected higher body weight among women who reported more time in caregiving and sedentary occupational activities as compared with women with lower body weights. However, this was not observed in this study as correlations between body weight and time spent in caregiving and sedentary occupational activities were low and not significant (data not shown). When the KPAS indexes that reflected moderate and vigorous types of activity, such as sports and exercise, were compared with Caltrac MET-min·d−1 scores, associations were on the order of r = 0.34 to 0.57. Correlations with the KPAS 3-point and 4-point summary indexes were on the order of 0.50 when expressed as MET-min·d−1, yet low and not significant when expressed as kcal·d−1. Although the Caltrac accelerometer is not without limitations, its value as a criterion measure of physical activity is two-fold. The Caltrac accelerometer is an objective measure of physical activity and sensitive to much of the miscellaneous movement performed thorough out the day. Still, it does not detect physical activity performed while stationary in a standing or sitting position. Thus, the Caltrac may not be a suitable criterion measure for physical activity that does not involve vertical accelerations. Because the Caltrac scores represent total (i.e., 24-h) physical activity, including sleep, activities not listed in the KPAS, and activities that do not register on the Caltrac accelerometer, lower correlations with specific types of activity are expected (14).
The indirect measures of physical activity, O2 peak, and percent body fat reflect habitual, moderate, and vigorous intensity physical activity. Correlations between the KPAS sports/exercise and active living habits indexes were positively related with O2 peak and inversely related with percent body fat. The KPAS housework/caregiving and occupation indexes reflect mostly light-to-moderate intensity activities (≤4 METs) (4,12,19). As expected, the correlations between the KPAS housework/caregiving activities and measures of O2 peak and percent body fat were low and not significant and are probably insufficient to modify cardiorespiratory fitness and body fat among regularly active individuals.
Comparison of the KPAS items and summary indexes with PA records allows direct comparison between survey responses and actual participation in similar activities. According to Sternfeld et al. (19), the rationale for developing the KPAS was to include gender-relevant items in a physical activity survey that reflect activities done in women’s lives. This is consistent with the conclusions from an expert panel meeting that identified important issues related to measuring physical activity in women, such as including gender-specific activities in physical activity surveys (9). In the KPAS, the housework/caregiving section includes child care, elder care, cooking, routine cleaning, garden and yard work, and home repair activities. Correlations for these items were moderate to high and statistically significant (r = 0.28 to 0.64). Higher correlations were observed among housework/caregiving activities that were performed daily or every other day. Correlations for infrequent activities, like major cleaning, grocery shopping, and outdoor house and yard work, were not statistically significant. However, the lack of statistical association between these activities may be more related to participants not doing these activities on days they kept their PA record rather than a true lack of association between the survey items and participation patterns among the study participants. Because none of the women in the comparison study reported doing elder care activities, this KPAS item could not be evaluated.
A criticism of physical activity surveys that focus only on sports and recreational items is that they can underestimate true activity patterns of nonsporting women (3,9). Studies among women show that housework and caregiving activities are a significant part of women’s lives. Shaw (15–17) presents data from Canadian studies using time-budget diaries to show that women spend a significant proportion of their day doing housework and caregiving tasks. Using in depth interviews among Indo Canadian women, Tirone and Shaw (21) reported that children, husbands, and extended family members were very important in women’s lives. The study participants reported that their extended families were a source of friendship and social support. The importance of housework activities in physical activity surveys was identified by Ainsworth et al. (3) in the Survey of Activity, Fitness, and Exercise study. Using a 4-wk history recall of physical activity, the investigators showed women’s daily energy expenditure exceeded men’s by 22% when housework activities were included in the physical activity score. However, when housework activities were omitted, men expended 27% more energy per day as compared with women. This was consistent with the greater amount of household activity reported by these women versus men in detailed physical activity records.
Exclusion of housework and caregiving activities from physical activity surveys can also convey a message to women that their participation in these activities is unimportant (9). Among the women in the KPAS evaluation study, housework and caregiving activities accounted for the second largest median energy expenditure (after occupational activity) of 356 MET-min·d−1. The PA record three-item summary score for sports/exercise, active living habits, and occupation was 800 MET-min·d−1. Addition of the housework/caregiving summary score increased the energy expenditure to 1260 MET-min·d−1. The additional information obtained about women’s housework/caregiving activities may be important in physical activity counseling settings or intervention studies designed to promote regular physical activity among women.
Some items from the KPAS were difficult to evaluate with the criterion measures of physical activity used in this study. We could not accurately evaluate survey items that asked about one’s general impression of their physical activity status as compared with others, or if they were physically tired at the end of the work day. KPAS indexes that included data not collected in the PA records (i.e., intensity ratings for occupation census codes and ratings of sweating at work and during sports/exercise) could not be optimally evaluated in the current comparison study. However, it is unlikely that the inability to evaluate these questions strongly biased the physical activity summary scores since most correlations between the KPAS and PA record summary scores were of substantial magnitude (7).
Concerns may arise with the choice of direct and indirect measures of physical activity selected for comparison with the survey items and indexes. Physical activity records were used to identify the type, duration, and perceived intensity of activities performed by the subjects during a 2-wk period. Although not a gold standard for physical activity, the PA records are an acceptable method to record physical activity (10). In this study, the Compendium of Physical Activity (4) was used to obtain MET estimates for the activities recorded in the PA records and the three most frequent sports and exercises recalled on the KPAS. The MET levels for most, but not all, activities in the Compendium of Physical Activity were obtained from laboratory and field measures that measured oxygen uptake using indirect calorimetry methods (4). For other activities, where measured MET levels were not available, MET levels were estimated from similar activities that had measured MET levels. Using the same MET intensities printed in the Compendium of Physical Activities to estimate MET-min·d−1 in the PA record and the KPAS may be a limitation to the present findings because the MET levels may differ among individuals for the same activity depending on their efficiency and speed of movement. Similar concerns can be expressed with the use of cardiorespiratory fitness and percent body fat as criterion measures of physical activity. Maximal cardiorespiratory fitness and percent body fat are influenced by many factors beside physical activity, including genetic predisposition, environmental conditions, measurement methods and technical errors, nutritional status, and test-retest variability (18,20,22,23). Thus, the correlation coefficients observed in this study between survey indexes and cardiorespiratory fitness and percent body fat should be interpreted in light of these considerations.
Finally, these results should be interpreted in relation to the sample from which they were obtained. Women in the KPAS evaluation study were volunteers who were interested in knowing about their cardiorespiratory fitness, body fatness, and general physical activity habits. Most women were white, between the ages of 30 and 40 yr, college educated, employed, and exhibited moderately good fitness levels, and above average body fat. It is unknown whether the results obtained in the current study sample would be observed in studies of women in lower SES settings, living in diverse areas of the country or the world, or among women from varied race/ethnicity as those measured in this study. Participants also varied in the types, frequency, and duration of their daily activities, ranging from sedentary to competitive athlete status. Overall, the sample may have been more active than national surveys of similar age women. The median energy expended in activity living habits was 63.7 MET-min·d−1 and in sports/exercise activities was 171 MET-min·d−1. Given an average body mass of 60 kg, energy expended in sports and exercise at these levels is sufficient to meet the 1996 Surgeon General’s guidelines of 150 kcal·d−1 in moderate and vigorous activities (22).
In summary, the KPAS has four physical activity indexes that alone, or in combination, explain from 12 to 53% of the variance in direct and indirect measures of physical activity. In this cross-sectional comparison study, we conclude that the KPAS identifies specific, habitual, daily activities, such as housework/caregiving, occupation, and sports/exercise activities, relatively well. More general and infrequent activities, as seen in some housework activities and in the active living index, correlated less well with specific activities recorded in a PA record. However, they were effective in characterizing general and habitual activity habits as reflected by measures of cardiorespiratory fitness, percent body fat, and the Caltrac accelerometer. Thus, the KPAS may be preferable for studies that try to characterize women’s routine, daily activities. Future research should compare the KPAS with criteria measures obtained over a longer period, such as 1 yr. This would provide further evaluation of the ability of this instrument to assess habitual levels of physical activity.
We thank Jeannie Ward Bensfield, Steven Criscoe, and Robert McMurray for their assistance with the study. Housework/Childcare data from this paper were used for the MA Thesis completed by Jeannie W. Bensfield in the Department of Physical Education, Exercise, and Sport Science at the University of North Carolina at Chapel Hill.
1. Ainsworth, B. E., D. R. Jacobs Jr., and A. S Leon. Validity and reliability of self-reported physical activity status: the Lipid Research Clinics Questionnaire
. Med. Sci. Sports Exerc. 25:92–98, 1993.
2. Ainsworth, B. E., H. L. Montoye, and A. S Leon. Methods of assessing physical activity during leisure
and at work. In:Physical Activity, Fitness, and Health: International Proceedings and Consensus Statement,
C. Bouchard, R. J. Shephard, and T. Stephens (Eds.). Champaign, IL: Human Kinetics, 1994, pp. 146–159.
3. Ainsworth, B. E., M. T. Richardson, A. S. Leon, and D. R. Jacobs Jr. Gender differences in self-reported physical activity. Women Sport Phys. Activity J. 23:1–16, 1993.
4. Ainsworth, B. E., W. L. Haskell, A. S. Leon, et al. Compendium of Physical Activities: classification of energy costs of human physical activities. Med. Sci. Sports Exerc. 25:71–80, 1993.
5. American College of Sports Medicine. Guidelines for Exercise
Testing and Prescription, 4th Ed. Philadelphia: Lea & Febiger, 1991, pp. 1–92.
6. Baecke, J. A. H., J. Burema, and J. E. R. Fritjers. A short questionnaire
for the measurement of habitual physical activity in epidemiological studies. Am. J. Clin. Nutr. 36:936–42, 1982.
7. Jacobs, D. R. Jr., B. E. Ainsworth, T. J. Hartman, and A. S. Leon. A simultaneous evaluation of ten commonly used physical activity questionnaires. Med. Sci. Sports Exerc. 25:81–91, 1993.
8. Kelsey J. L., W. D. Thompson, and A. S. Evans. Methods in Observational Epidemiology. New York: Oxford University Press, 1986, pp. 290–292.
9. Masse, L. C., B. E. Ainsworth, S. Tortolero, et al. Measuring physical activity in midlife, older, and minority women: issues from an expert panel. J. Women’s Health 7:57–67, 1998.
10. Montoye, H. J., H. C. G. Kemper, W. H. M. Saris, and R. A. Washburn. Measuring Physical Activity and Energy Expenditure. Champaign, IL: Human Kinetics, 1996, pp. 1–183.
11. Mifflin, M. D., S. T. St. Jeor, L. A. Hill, B. J. Scott, S. A. Daugherty, and Y. O. Koh. A new predictive equation for resting energy expenditure in healthy individuals. Am. J. Clin. Nutr. 51:241–247, 1990.
12. Pate, R. R., M. Pratt, S. N. Blair, et al. Physical activity and public health: a recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. JAMA 273:402–407, 1995.
13. Pereira, M. A., S. J. Fitzgerald, E. W. Gregg, et al. A collection of physical activity questionnaires for health related research. Med. Sci. Sports Exerc. 29:S1–S205, 1997.
14. Richardson, M. T., A. S. Leon, D. R. Jacobs Jr., B. E. Ainsworth, and R. C. Serfass. Ability of the Caltrac accelerometer to assess daily physical activity levels. J. Cardiopulm. Rehabil. 15:107–113, 1995.
15. Shaw, S. M. Leisure
, or free time? Measuring time usage. J. Leisure
Res. 18:177–189, 1986.
16. Shaw, S. M. Gender and leisure
: inequality in the distribution of leisure
time. J. Leisure
Res. 17:266–282, 1985.
17. Shaw, S. M. Women’s leisure
time: using time budget data to examine current trends and future predictions. Leisure
Stud. 10:171–181, 1991.
18. Siri, W. The gross composition of the body. Adv. Biol. Med. Phys. 4:239–272, 1956.
19. Sternfeld, B., B. E. Ainsworth, and C. P. Quesenberry Jr. Physical activity patterns in a diverse population of women. Prev. Med. 28:313–323, 1999.
20. Taylor, H. L., E. Buskirk, and A. Henschel. Maximal oxygen intake as an objective measure of cardiorespiratory performance. J. Appl. Physiol. 8:73–80, 1955.
21. Tirone, S. C., and S. M. Shaw. At the center of their lives: Indo Canadian women, their families and leisure
. J. Leisure
Res. 29:225–244, 1997.
22. U. S. Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996, pp. 1–259.
23. Wilmore, J. H. A simplified method for determination of residual lung volumes. J. Appl. Physiol. 27:96–100, 1969.
Keywords:© 2000 Lippincott Williams & Wilkins, Inc.
PHYSICAL FITNESS; EXERCISE; RECREATION; LEISURE; QUESTIONNAIRE