The high prevalence of overweight and obesity affects Americans' health and the economic well-being of the United States. Approximately two-thirds of American adults are either overweight or obese (1). Excess weight increases risk for cardiovascular diseases, cancer, and premature death (2). A recent analysis indicates that obesity accounted for more than 100,000 excess deaths in 2000 (3). The economic cost of obesity in 1995 was $99.2 billion, representing 5.7% of US health expenditures for that year (4).

Physical activity contributes to weight loss and maintenance. Although several groups recommend 30 minutes of daily moderate physical activity (MPA) (5,6), more may be needed to promote weight loss and weight maintenance. An Institute of Medicine report recommends at least 60 minutes of daily MPA to maintain a healthy weight (7). Although researchers know that overweight and obese people exercise less than do their healthy-weight peers (8), little is known about the level of activity and how it compares with recommendations.

Physical activity is typically assessed from self-report. Recent technological advances have made accelerometry, a way to objectively assess physical activity, available for large-scale research trials. Objective measures overcome many of the limitations of self-report instruments, particularly measurement error associated with recall (9). In addition, obese people may overreport physical activity (10), which increases recall bias. Accelerometers can more accurately assess intensity, duration, and daily patterns of physical activity without relying on error-prone recall methods.

We used accelerometers to measure baseline physical activity in a large sample of overweight and obese people who were screened for participation in the Weight Loss Maintenance Trial. We measured total minutes of MPA, vigorous physical activity (VPA), and moderate-to-vigorous physical activity (MVPA) and activity performed in bouts of at least 10 minutes (the minimum amount associated with health benefits [6]). This information is needed to more effectively promote physical activity among overweight and obese people, who are at excess risk for chronic disease and premature death.

Of the 1,685 participants who were screened for inclusion in the Weight Loss Maintenance Trial, 1,648 (98%) wore their accelerometers for at least 6 hours per day and made up the study sample. Of the 1,685 who were screened, 82% wore their accelerometers for at least 10 hours per day. Mean age of participants was 60 years (standard deviation, 9 years). Two-thirds of participants were women, 43% were African American, and 79% were obese. Mean BMI was 28.2 kg/m² among overweight participants and 35.9 kg/m² among obese participants.

On average, participants were active for less than 16 minutes per day, and less than 3 of these minutes was spent in VPA (Table 1). Participants who were younger, male, non-African American, and overweight were significantly more active than were participants who were older, female, African American, and obese.

Although mean minutes of MVPA differed significantly by weekday vs weekend day, the magnitude of the difference was negligible (0.2 min) (Table 2). Overall, participants were more active in the mornings than in the afternoons and least active in the evenings. Although subgroup data were not tested for significance, this trend was observed for all demographic and weight status subgroups.

A total of 810 (49%) participants recorded at least 1 MVPA bout of at least 10 minutes during the monitoring period. These bouts lasted an average of 33.3 minutes (Table 3) and lasted significantly longer for men and overweight participants than for women and obese participants, although the difference by weight status was minimal (0.3 minutes). Bouts lasted approximately 5 minutes longer on weekend days than on weekdays and approximately 5 minutes longer in the morning than in the afternoon or evening. Participants were most active (in terms of total MVPA minutes as well as bout minutes) on weekend mornings; in comparison, participants recorded 7 to 8 fewer minutes of MVPA on weekday and weekend evenings.

When all MVPA minutes were considered, approximately one-fourth of participants met recommendations of 30 minutes of MVPA on 5 days per week (Table 4). Forty percent of men and 36% of overweight participants met the recommendation of at least 150 minutes of MVPA in 7 days. In contrast, only 18% of women and 23% of obese participants met this recommendation. Few participants (≈2%) met the more stringent target of at least 60 minutes of MVPA per day.

On average, our participants were active for less than 16 minutes per day. Almost half of participants had at least 1 bout of MVPA that lasted at least 10 minutes, and each bout, on average, exceeded 30 minutes. MVPA was most likely in the morning and on weekends, and mornings were times in which MVPA was most likely to occur.

Similar findings have been reported in national data sets that used self-report instruments (17,18). A previous study used accelerometry data from the National Health and Nutrition Examination Survey (NHANES) and found similar age, sex, and race associations with physical activity, although results were not reported by weight status (19). Another study that used accelerometers showed that overweight adults were more physically active than were obese adults, but the sample size was small and mean differences were not significant (20). In our large, geographically diverse sample, obese adults were significantly less active than were their overweight peers. However, because our study design was cross-sectional, we cannot determine if obesity decreased physical activity or if lower physical activity over time contributed to obesity.

This study is among the first large-scale studies to objectively measure physical activity and examine activity levels across different times of day. In general, overweight and obese adults were more likely to be active in the morning, particularly on weekends. Adults aged 50 years or more were active 2 minutes less than were adults aged less than 50 years, and this finding supports previous findings that older adults are less active in the evening than are younger adults (21). Another study reported more activity on weekends than on weekdays (8), which is consistent with our results. These results suggest that studies of physical activity patterns should examine time of day and weekdays vs weekend.

Information from this study has implications for physical activity promotion. Overweight and obese adults were most active on weekends and mornings. Efforts to encourage additional physical activity at times when adults are most likely to be active should promote weekend mornings. Conversely, efforts that seek to encourage activity at times when baseline activity is low should promote weekday afternoons and evenings.

Because of the various methods in which physical activity data are collected and reported, making comparisons across studies is difficult. National self-report data suggest that approximately 50% of all adults (17) and 20% to 30% of overweight and obese adults (22,23) meet national recommendations. National data on minutes of MVPA are only available from NHANES accelerometry data. On the basis of the mean number of minutes reported, our sample was less active than was the general US population (19). Physical activity assessed with accelerometry can provide more precise information regarding MVPA minutes than is available from self-report.

NHANES accelerometry data indicate that only 3.5% of US adults meet physical activity recommendations (19). This percentage contrasts with the more than 25% of our participants who met recommendations. The difference is surprising because our participants were all overweight or obese, conditions associated with low physical activity. Methodological differences may explain the discrepancy. In the NHANES data, investigators only considered bouts of at least 10 minutes, which had to total 30 minutes per day on 5 days per week. We counted all minutes of MVPA and considered 150 minutes in 7 days to meet recommendations. Given that less than half of our sample engaged in even 1 bout of at least 10 minutes, if we had used more conservative criteria for meeting recommendations, our findings probably would have reflected those of the NHANES data.

Few participants met the more stringent recommendation of at least 60 minutes of physical activity per day. A recent study (8) reported that 13% of overweight and obese participants met this recommendation, but that sample was smaller than ours (n = 62), and participants were specifically recruited into a study that evaluated exercise and energy expenditure, which may have attracted more active participants.

The number of minutes of physical activity determined by accelerometry depends on the cutpoints used, which could overestimate or underestimate true physical activity levels. Only 1 validation study used the RT3 to assess accelerometry counts by oxygen uptake (14). Additional validation studies are needed to determine appropriate cutpoints for light, moderate, and vigorous physical activity for instruments such as the RT3.

Physical activity data are often skewed, and this was the case with our data. We considered presenting the data as medians and interquartile ranges rather than as means and standard deviations. However, other studies of physical activity that used accelerometry reported means (8,19), and we chose this approach to be consistent with other published work.

Strengths of our study are a large sample size, the use of accelerometry data, high adherence with wearing the accelerometer (98%), diverse demographic characteristics of the sample, and our ability to examine daily and weekly patterns of physical activity. Although NHANES collected accelerometer data from 2002 through 2004 (19), this is the first report that focuses solely on overweight and obese adults. The major limitation of these data is that participants were enrolling in a weight loss trial, and these findings may not be entirely generalizable to the US overweight and obese population.

We found that previously reported differences in physical activity by demographic features and weight status persist when physical activity is measured with an objective method. We also found that obese people are physically active, although not at recommended levels. Weekend mornings, a time when our participants were more active, may be an opportune time to promote additional physical activity, particularly in bouts of 10 minutes or longer, to meet national recommendations.