On the basis of recent estimates, approximately 7.8% of the population in the United States has type 2 diabetes mellitus, including 17.9 million diagnosed and 5.7 undiagnosed cases (10). Moreover, when using an age criterion of ≥20 yr or ≥60 yr, the prevalence is estimated to be 10.7% and 23.1%, respectively (10). This creates a significant public health burden because of the direct cost of its treatment, the indirect cost associated with it, and its link to increased mortality due to cardiovascular disease (CVD) (10,16,18).
The recent Physical Activity Guidelines Advisory Committee Report concluded that there is evidence that physical activity is important in the treatment of type 2 diabetes because it improves glycemic and cardiovascular risk factor control (21). For example, a meta-analysis conducted by Thomas et al. (19) concluded that repeated bouts of physical activity improve glycemic control in persons with type 2 diabetes. Observational cohort studies have associated higher levels of walking and leisure time physical activity (LTPA) with reduced CVD incidence and mortality (5).
The vast majority of studies examining the influence of physical activity on diabetes treatment and related complications have been based on self-report with limited objective data on the daily patterns of physical activity. An objective measure of physical activity using accelerometry was included in the multicenter Look AHEAD Study on individuals with type 2 diabetes from 8 of the 16 clinical sites (14). The availability of these data provides an opportunity to objectively examine patterns of physical activity in this population and to determine factors that may influence participation in physical activity.
The purpose of this study was to describe the patterns and correlates of physical activity in individuals with type 2 diabetes mellitus enrolled in the multicenter Look AHEAD Study using an objective measure of physical activity (accelerometry). Specifically, we examine whether patterns of physical activity among persons with type 2 diabetes vary according to levels of body mass index (BMI), sex, ethnicity/race, age, baseline fitness, diabetes medication, and presence of CVD.
Participants for this study are a subgroup of participants who completed baseline testing before being randomized into the multicenter Look AHEAD Study, which is a trial examining the effect of weight loss on the primary and secondary prevention of CVD in obese adults with type 2 diabetes (14). The Look AHEAD Study included 5145 participants recruited across 16 clinical sites in the United States, with 8 of the sites including 2627 participants enrolled in a substudy to assess physical activity using accelerometry. Descriptive data for this subgroup of participants are presented in Table 1.
All participants gave written informed consent approved by the local institutional review boards before participating in this study. Participants completed initial eligibility assessments and provided baseline data before randomization, and these data are used for the analyses presented in this article. These data include demographic characteristics (age, sex, race/ethnicity, CVD history (self-report of myocardial infarction, stroke, transient ischemic event, percutaneous transluminal coronary angioplasty, or coronary artery bypass graft), use of diabetes medication), physical activity assessed using accelerometry, physical activity assessed using self-report, and cardiorespiratory fitness according to a standardized protocol as described in the next paragraphs.
Assessment of Physical Activity
The RT3 triaxial accelerometer (StayHealthy®, Monrovia, CA) was used to provide an objective measure of physical activity. On the basis of the recommendations from the manufacturer, participants were instructed to wear the device at a location consistent with the anatomical location of the anterior iliac spine with the accelerometer placed in a vertical position. The accelerometer was worn on the belt or waistband of the clothing; however, participants were also provided with a belt they could wear to secure the accelerometer to the proper location at the waist if it could not be attached properly to their clothing. Participants were also provided written instructions that included the proper placement of the accelerometer on the body, specific dates (d) that the accelerometer should be worn, procedures to follow should there be difficulty with the operation of the accelerometer, and procedures for returning the accelerometer. The data collection mode for the accelerometer was set in the three-axis and 1-min epoch mode.
Participants were provided the accelerometer during a visit to the research center as part of an assessment visit, and they were requested to wear the accelerometer for the remainder of that day (partial day) and for six consecutive days after the visit (six full days) and to wear it until it was returned to the clinic on the seventh day. Thus, the protocol allowed for up to six full days of data collection. Participants were instructed to wear the accelerometer during all waking hours during this period and to remove it only during periods of bathing, showering, or other water-based activities. Quality control procedures were implemented, and if data were deemed to be invalid, an attempt was made to have the subject wear the accelerometer for an additional period to provide valid data.
Data for a given day were considered valid if the accelerometer was worn for ≥10 h on that day, which is consistent with previously published recommendations for the use of an accelerometer to measure physical activity (9). The accelerometer was determined not to have been worn for periods defined as ≥20 continuous minutes of no activity counts on the accelerometer, and this was subtracted from 24 h (1440 min) to determine whether the ≥10-h·d−1 criterion was achieved so that a day could be considered a valid one. To be included in the analyses, a subject had to provide at least four valid days of accelerometry data.
Total energy expenditure per minute (kcal·min−1) and estimated resting energy expenditure (kcal·min−1) were provided by the StayHealthy® software that accompanied the RT3 accelerometer. Using these data, METs per minute were computed by dividing the estimated total energy expenditure per minute by the estimated resting energy expenditure (METs = total energy expenditure per minute/estimated resting energy expenditure). Physical activity periods that are consistent with public health recommendations were identified. The American College of Sports Medicine and the American Heart Association define moderate-intensity physical activity as ≥3 MET·min−1, and it has been recommended that bouts of activity be at ≥10 min in duration (6). Thus, accelerometry data were analyzed to identify periods that were consistent with these recommendations (≥3 MET·min−1 and ≥10 min in duration). Vigorous bouts of physical activity consistent with ≥6 MET·min−1 and ≥10 min in duration were also identified for analysis.
Within the Look AHEAD Study, LTPA was assessed on a subsample of approximately 50% of the participants at baseline, which reflects the same subject pool as those who completed measurements using accelerometry. Self-reported LTPA was assessed using the questionnaire developed for the Harvard Alumni Study (12) as part of a structured interview. LTPA was assessed by participants reporting their activity during the past week. Activities assessed included flights of stairs (10 steps = 1 flight), number of city blocks walked (1 mile = 12 blocks), and other sport, recreational, or fitness activities. Standardized scoring procedures were used to compute energy expenditure (kcal·wk−1) of LTPA (12). Participants also reported the number of days per week that they participated in LTPA that was of sufficient intensity to "engage in regular activity akin to brisk walking, jogging, bicycling, etc. long enough to work up a sweat, get your heart thumping, or get out of breath," which will be called "sweat episodes."
Additional Baseline Assessments
A questionnaire was used to assess age, sex, ethnicity, current diabetes treatment, and history of CVD. Previously described methods were used to measure height, weight, and BMI (8). A maximal graded exercise treadmill test was used to assess cardiorespiratory fitness at baseline as previously described (22). This test was terminated at the point of volitional exhaustion or at the point where the American College of Sports Medicine (1) test termination criteria were observed. A baseline test was considered valid if the maximal HR was ≥85% of age-predicted maximal heart rate (HRmax = 220 − age) if the participant was not taking a β-adrenergic blocking medication (β-blocker). If the participant was taking a β-blocking medicine, the baseline test was considered valid if RPE was ≥18 at the point of termination. Moreover, to be eligible for participation in the Look AHEAD Study, the participant needed to achieve ≥4 METs on the baseline graded exercise test, where 1 MET is equal to 3.5 mL·kg−1·min−1 of oxygen uptake.
All analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC). Statistical significance was defined as P ≤ 0.05. Bivariate analyses were performed to examine differences in bouts per day meeting the criteria of ≥10 min per bout and either ≥3 MET·min−1 or ≥6 MET·min−1 within categories of BMI, sex, race/ethnicity, age, fitness, diabetes medication usage, and history of CVD. For individuals found to have at least one bout of physical activity meeting the criteria of either ≥10 min per bout or ≥3 MET·min−1 or either ≥10 min per bout or ≥6 MET·min−1, additional bivariate analyses were performed to test for differences in minutes per bout, METs per minute, and MET-minutes for all bouts meeting the criteria. Differences in self-reported LTPA (kcal·wk−1 and sweat episodes per week) within categories of BMI, sex, race/ethnicity, age, fitness, diabetes medication usage, and history of CVD were examined.
Multivariate analyses were performed using PROC GLM in SAS to determine the effect of age, BMI, sex, race/ethnicity, fitness, diabetes medication, and history of CVD (self-report of myocardial infarction, stroke, transient ischemic event, percutaneous transluminal coronary angioplasty, or coronary artery bypass graft) on objective measures of physical activity using accelerometry. Analyses were performed with bouts per day, minutes per bout, METs per minute, and MET-minutes considered as the dependent measures. Separate analyses were performed for these dependent measures for bouts of activity meeting the criteria of ≥10 min per bout and either ≥3 METs per bout or ≥6 METs per bout.
Of the 5145 participants randomized into the Look AHEAD Study, 2627 were at the eight Look AHEAD clinical sites participating in the accelerometry substudy. Of these participants, 2240 (85.3% of the 2627 participants) provided accelerometry data at baseline, with reasons for missing data including equipment failure, participant refusal, time/scheduling, and other reasons. Within the "other reasons" category, some clinics did not consider this portion of the Look AHEAD trial as part of the larger trial, and thus, not all participants at these clinic sites were presented with an opportunity to consent and participate in the study, with this reason accounting for approximately 21% of the missing baseline accelerometry files. Visual inspection of individual data graphs was performed to determine the presence of activity energy expenditure as depicted by vertical lines during the wear period, the presence of episodes of activity interspersed with sustained periods of zero activity indicating that the unit was removed for sleeping, and the weekly activity energy expenditure as within reasonable limits for this population (between 200 and 15,000 kcal). Using these criteria, 95.8% (n = 2145) of the data files available were valid and were considered for analysis in this study. Baseline demographic data are presented in Table 1.
Separate analyses were performed using all days that achieved the minimum of ≥10 h·d−1 of wear time, along with the first 4, 5, or 6 days that achieved the minimum of ≥10 h·d−1 of wear time. The pattern of results did not change when data were analyzed using 4, 5, 6, or all days that met the ≥10 h·d−1 of wear time (data not shown), and therefore, the following results are described using data from all days that met this criterion of ≥10 h of wear time.
Comparison by BMI categories.
Participants were categorized on the basis of baseline BMI as 25.0 to <27.0, 27.0 to <30.0, 30.0 to <35.0, 35.0 to <40.0, and ≥40 kg·m−2 (Tables 2 and 3). There was a significant difference between BMI categories for bouts per day achieving the ≥3-METs and ≥10-min duration criteria, with fewer bouts identified in those with greater BMI. For participants who had a period meeting these criteria, the duration per bout was lower at higher BMI categories. A similar pattern was shown for METs per bout, MET-minutes for bouts meeting these criteria, and bouts of activity per day. Individuals at higher categories of BMI also had fewer bouts per day, had shorter duration of activity bouts, had lower METs per bout, and had lower MET-minutes when the criteria of ≥6 METs and ≥10 min were applied to the data.
Comparison by sex.
There were significantly more bouts per day in men than in women when using the ≥3-METs and ≥10-min duration criteria (0.83 ± 1.00 vs 0.45 ± 0.73 bouts per day, P < 0.0001; Table 2). Moreover, for bouts identified that met these criteria, duration per bout was significantly longer for men (20.65 ± 10.26 min) than women (18.63 ± 8.64 min, P < 0.0001). METs per minute for these bouts did not differ by sex; however, total MET-minutes for all bouts meeting the criteria of ≥3-METs and ≥10-min duration were higher in men than in women (P = 0.0005).
When the criteria of ≥6 METs and ≥10 min were applied to the data, men continued to have more activity bouts that met these criteria than women have (0.16 ± 0.35 vs 0.09 ± 0.28 bouts per day, P < 0.0001; Table 3). The duration per bout did not differ between sexes (18.61 ± 8.71 vs 18.83 ± 9.11 min), but METs per minute were greater for men (8.05 ± 1.18 METs) compared with those for women (7.58 ± 1.04 METs, P < 0.0001). However, total MET-minutes for all bouts meeting the criteria of ≥3-METs and ≥10-min duration did not differ by sex.
Comparison by ethnic/racial group.
There was a significant effect of ethnic/racial group on bouts per day and minutes per bout when using the criteria of ≥3 METs with ≥10-min duration, with individuals identified as African-American/black having the fewest bouts per day and the lowest minutes per bout (Table 2). African-American/black individuals also had the fewest bouts per day when the criteria of ≥6 METs with ≥10-min duration were applied to the data (Table 3). However, there was no significant effect of ethnicity/race for duration per bout, METs per minute, or total MET-minutes using either of these criteria.
Comparison by age.
There was no significant difference between age categories for bouts per day of participating in activity ≥10 min in duration when defined using either the ≥3-METs or the ≥6-METs category. However, METs per minute decreased with advancing age for ≥3-METs activities, whereas minutes per bout and total MET-minutes decreased with advancing age for ≥6-METs activities (Tables 2 and 3).
Comparison by baseline fitness.
Bouts per day for both ≥3 and ≥6 METs were positively associated with fitness. For individuals reporting participation in bouts that were ≥3 or ≥6 METs that were ≥10 min in duration, there was a significant increase in duration per bout, METs per minute, and total MET-minutes of physical activity meeting these criteria with increasing levels of fitness (Tables 2 and 3).
Comparison by type of diabetes treatment and by history of CVD.
There was no significant association between type of diabetes medication and bouts per day, duration per bout, METs per minute, or total MET-minutes for either the ≥3-METs or ≥6-METs physical activity category. When grouped according to history of CVD, individuals who had a history of CVD and participated in activity bouts of ≥3 METs that were of at least 10 min in duration had higher METs per minute (5.17 ± 1.01 METs) than did those without a history of CVD (5.03 ± 0.98 METs, P = 0.0463). Data are presented in Tables 2 and 3.
Self-reported LTPA was compared across categories of BMI, sex, race, age, fitness, diabetes medication usage, and history of CVD (Table 4). Energy expenditure in LTPA was negatively associated with BMI, was greater in men than women, was lowest in individuals classified as African-American/black, and was positively associated with fitness. There was no effect of age, type of diabetes medication usage, or history of CVD on energy expenditure in self-reported LTPA. A similar pattern of results was observed when data were analyzed for "sweat episodes per week."
Results of the multivariate models using the criteria of ≥10 min in duration and ≥3 METs to examine components of physical activity are shown in Table 5. Bouts per day were significantly higher in men than in women and were positively associated with fitness (P < 0.0001). Minutes per bout were significantly higher with increasing levels of fitness (P ≤ 0.0001), with a trend for fewer minutes per bout with increasing BMI (P = 0.0572). METs per minute were significantly higher in women than in men (P = 0.0001) and were positively associated with fitness (P < 0.0001), with a trend observed for increasing METs per minute with BMI (P < 0.0982). However, total MET-minutes for bouts meeting the criteria of ≥10 min in duration and ≥3 METs were positively associated with fitness (P < 0.0001), with a trend for a positive association between total MET-minutes for bouts meeting the criteria and age (P = 0.0821).
Results of the multivariate models using the criteria of ≥10 min in duration and ≥6 METs to examine components of physical activity are shown in Table 6. Bouts per day were positively associated with age (P = 0.0099) and fitness (P < 0.0001) and were influenced by race/ethnicity, with African-American/black and other/mixed/missing having lower values that whites. Minutes per bout were significantly higher with increasing levels of fitness (P ≤ 0.0001). METs per minute were negatively associated with BMI (P = 0.0327), were lower in women than in men (P = 0.006), and were positively associated with fitness (P < 0.0001), with a trend observed for increasing METs per minute with BMI (P < 0.0982). However, total MET-minutes for bouts meeting the criteria of ≥10 min in duration and ≥3 METs were positively associated with fitness (P < 0.0001).
The most recent Physical Activity Guidelines for Americans have recommended at least 150 min·wk−1 of moderate-intensity activity that is spread throughout the week (20). Adults with type 2 diabetes in the Look AHEAD Study are participating in less than one bout per day of physical activity that meet the criteria of ≥3 METs and ≥10 min in duration on the basis of accelerometry data (Table 2), <0.20 bouts per day of vigorous-intensity activity (≥6 METs; Table 3) and approximately two sweat episodes per week or fewer on the basis of self-reported LTPA (Table 4). On the basis of accelerometry, results showed that 71% of subjects had at least one activity bout that met the ≥3 METs and ≥10 min in duration criteria, with 26% meeting the ≥6 METs and ≥10 min in duration. For those engaged in at least one activity bout per week that meet either the moderate- to vigorous-intensity or the vigorous-intensity criteria for at least 10 consecutive minutes, the duration of these bouts of activity was approximately 15-20 min (Tables 2 and 3), suggesting that few individuals were achieving the recommended level of physical activity. Moreover, 51% of subjects reported at least one sweat episode per week on the basis of self-report in this study. Nwasuruba et al. (11) reported that between 23% and 37% of adults with type 2 diabetes were meeting the recommended level of physical activity on the basis of the 2003 Behavioral Risk Factor Surveillance Survey data. Thus, there seems to be a need for interventions to increase the frequency and duration of sufficient doses of physical activity in individuals with type 2 diabetes to meet the recommended physical activity guidelines for adults, and this may improve diabetes control and minimize the risk of health complications or mortality related to this condition.
The data presented in this article include objectively measured physical activity using accelerometry, which also provide information on the patterns of physical activity in individuals with type 2 diabetes enrolled in the Look AHEAD Study. We conducted extensive electronic library searches to identify other studies of individuals with type 2 diabetes that have used objective measures of physical activity and found few such studies. Samuel-Hodge et al. (15) used a uniaxial accelerometer to assess baseline levels of physical activity in an intervention study and reported on the minutes per day of light-, moderate-, and vigorous-intensity activities. However, this provides limited information on the pattern of activity that is moderate to vigorous in intensity, which also meets the minimum recommended duration of 10 min as stated by Haskell et al. (6) and by the recent Physical Activity Guidelines for Americans (20). Thus, baseline data presented here from the Look AHEAD Study define activity on the basis of the ability of a bout to meet the ≥10-min duration while also meeting moderate- to vigorous- or vigorous-intensity criterion. This may provide a clearer perspective of the patterns of physical activity in individuals with type 2 diabetes that meet the criteria for physical activity bouts that are recommended by leading health organizations compared with other studies that have examined physical activity in individuals with type 2 diabetes.
Despite the advantages of using objective monitoring, the use of accelerometry is not without limitations. For example, studies have regularly demonstrated that accelerometry seems to be best suited to provide valid and reliable data for activities that have a movement pattern similar to level walking and jogging (7). Moreover, although activity counts have been used to define moderate and vigorous levels of physical activity, these activity counts vary by type of accelerometer and mode of activity. For example, the activity count thresholds for moderate and vigorous physical activity differ for the RT3 accelerometer (13) compared with those for other accelerometers (4,17). Rowlands et al. (13) also reported that activity count thresholds for men when treadmill walking were 400 counts less than the activity count thresholds representing a variety of moderate to vigorous activities. Moreover, there are no published activity count thresholds for the RT3 accelerometer to define moderate or vigorous activity in overweight and obese older adults with type 2 diabetes who are the population group examined in this study. The activity count thresholds to define moderate and vigorous physical activity reported by Rowlands et al. (13) were developed on the basis of 19 boys with a mean age of 9.5 ± 0.8 yr and of 15 men with a mean age of 20.7 ± 1.4 yr. Therefore, we computed METs from the total energy expenditure and resting energy expenditure data as defined earlier to identify periods that met ≥3 or ≥6 METs, which are the criteria for moderate and vigorous levels of physical activity.
On the basis of data from the 2003 Behavioral Risk Factor Surveillance Survey, African-American/black adults with type 2 diabetes were less likely than white adults to engage in sufficient physical activity to meet the recommendation of at least 30 min·d−1 of moderate-intensity activity on 5 d·wk−1 or more or 20 min·d−1 of vigorous activity on 3 d·wk−1 or more (11). On the basis of bivariate analyses presented for moderate- to vigorous-intensity physical activity, defined as ≥3 METs and ≥10 min in duration, it seems that African-Americans/blacks participate in fewer bouts per day and fewer minutes per bout than whites with type 2 diabetes, and this may contribute to the lower total MET-minutes of physical activity (Table 2). For vigorous physical activity (≥3 METs and ≥10 min in duration), this study showed that African-Americans/blacks participants engaged in the fewest bouts of this level of physical activity (Table 3). The multivariate model seems to confirm a race/ethnicity effect for bouts per day of either moderate- to vigorous-intensity (≥3 METs) or vigorous-intensity (≥6 METs) physical activity, with African-Americans/blacks participating in fewer bouts per day of activity when compared with whites (Tables 5 and 6). A similar pattern was observed when examining self-reported LTPA (Table 4). Thus, although all individuals with type 2 diabetes should be encouraged to participate in physical activity as part of their comprehensive treatment plan, particular attention may need to be given to interventions that target African-American/blacks and other racial/ethnic groups who seem to have lower levels of physical activity.
The importance of understanding patterns of physical activity in this population is highlighted by the growing evidence that physical activity may be important for reducing mortality due to CVD or to all causes in persons with type 2 diabetes. The review by Casperson and Fulton (2) of studies examining the effect of walking on all-cause and CVD mortality concluded that the reduction in risk ranged from 40% to 55%. Specifically, Gregg et al. (5) reported that the relative risk of all-cause mortality was 0.61 and 0.46 with 2.0-2.9 and 3.0-3.9 h·wk−1 of walking compared with 0 h·wk−1, respectively. The risk of CVD was reduced by 53% with 3.0-3.9 h·wk−1 of walking. Tanasescu et al. (18) also reported a significant reduction in the risk of all-cause mortality of 40% with >5.3 h·wk−1 of walking in adults with type 2 diabetes. Smith et al. (16) also reported hazard ratios of 0.54 and 0.19 for all-cause and CVD mortalities, respectively, for adults with diabetes who walked ≥1 mile·d−1 compared with those who did not walk. Although these data consistently show the beneficial effect of physical activity on the reduction in mortality in individuals with type 2 diabetes, these data are based on self-report and provide little information on the patterns of physical activity. Understanding the patterns of physical activity may inform interventions that target increases in physical activity in patients with type 2 diabetes.
Church et al. (3) reported on the importance of fitness to reduce CVD mortality in adults with type 2 diabetes. There was a 1.2 hazard ratio for cardiovascular mortality for each 1-MET difference in fitness on the basis of a multivariate analysis. The data presented in this current analysis from the Look AHEAD Study provide information on the pattern of physical activity that may enhance fitness in individuals with type 2 diabetes. The data presented in Tables 2 and 3 show greater frequency (bouts per day), duration (minutes per bout), and intensity (MET·min−1) of both moderate- to vigorous-intensity (≥3 METs) and vigorous-intensity activity (≥6 METs) with greater levels of fitness, and this was confirmed by the multivariate analyses (Tables 5 and 6). Moreover, a similar pattern was shown with self-reported LTPA with energy expenditure, and sweat episodes per week were positively associated with fitness (Table 4). Thus, although cross-sectional, these data suggest that individuals with type 2 diabetes can improve their level of fitness through participation in sufficient doses of physical activity, and according to Church et al. (3), this may reduce the risk of mortality. The data presented in this current study provide insight into the dose and pattern of physical activity that will elicit improvements in fitness in overweight and obese adults with type 2 diabetes.
It is recognized that the R2 values for the multivariate models are relatively low (Tables 5 and 6). This would suggest that the demographic factors considered in this study are collectively having modest effects on the observed patterns of physical activity in this study. Therefore, there may be additional demographic, behavioral, or physiological factors that are influencing physical activity behaviors in this group of individuals with diabetes, and this will require further examination. Moreover, the Look AHEAD Study provides a unique opportunity to examine changes in physical activity patterns prospectively and to examine factors that influence observed changes in physical activity in adults with diabetes as they age. An interesting finding is that higher levels of physical activity were consistently associated with higher levels of fitness in multivariate models after controlling for other demographic characteristics. These data may suggest that, although patterns of physical activity may vary by demographic characteristics, it is important to engage in physical activity to improve fitness in patients with type 2 diabetes, and this may significantly influence the risk of mortality in this population (3).
In summary, these findings represent one of the few studies that have used objective measurement to examine the patterns of physical activity in individuals with type 2 diabetes. In this sample of more than 2000 participants in the Look AHEAD Study with diabetes, it does not seem that the vast majority of individuals meet the current recommended levels of physical activity (6,20). However, the patterns of activity vary consistently by level of fitness, with higher levels of fitness characterized by a higher frequency, duration, and intensity of activity bouts that are moderate to vigorous or vigorous in intensity. In multivariate analyses, age, BMI, type of diabetes medication, or history of CVD had little influence on the patterns of physical activity. However, sex and race/ethnicity may influence the pattern in individuals with type 2 diabetes, and this should be considered when developing physical activity interventions in this population. The Look AHEAD Study provides an opportunity to examine changes in physical activity measured objectively using accelerometry resulting from a lifestyle intervention, which can provide valuable information on the effect of activity patterns on weight loss and other health-related outcomes in individuals with type 2 diabetes.
Clinical trial registration number: NCT00017953, clinicaltrials.gov.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
Additional acknowledgments are provided in Supplemental Digital Content 1 (https://links.lww.com/MSS/A36).
1. American College of Sports Medicine. ACSM's Guidelines for Exercise Testing and Prescription
. Philadelphia (PA): Lippincott, Williams & Wilkins; 2010. p. 83.
2. Casperson CJ, Fulton JE. Epidemiology of walking and type 2 diabetes. Med Sci Sports Exerc
. 2008;40(7 suppl):S519-28.
3. Church TS, LaMonte MJ, Barlow CE, Blair SN. Cardiorespiratory fitness and body mass index as predictors of cardiovascular disease mortality among men with diabetes. Arch Intern Med
4. Freedson PS, Melanson E, Sirard J. Calibration of the Computer Science and Applications, Inc. accelerometer. Med Sci Sports Exerc
5. Gregg EW, Gerzoff GL, Casperson CJ, Williamson DF, Narayan KMV. Relationship of walking to mortality among US adults with diabetes. Arch Int Med
6. Haskell WL, Lee I-M, Pate RR, et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc
7. Jakicic JM, Winters C, Lagally K, Ho J, Robertson RJ, Wing RR. The accuracy of the TriTrac-R3D accelerometer to estimate energy expenditure. Med Sci Sports Exerc
8. Look AHEAD Research Group. Reduction in weight and cardiovascular disease risk factors in individuals with type 2 diabetes: one-year results of the Look AHEAD trial. Diabetes Care
9. Masse LC, Fuemmeler BF, Anderson CB, et al. Accelerometer data reduction: a comparison of four reduction algorithms on select outcome variables. Med Sci Sports Exerc
. 2005;37(11 suppl):S544-54.
10. National Diabetes Information Clearinghouse (NDCI) Web site [Internet]. Washington (DC): US Department of Health and Human Services; [cited 2009 Jan 19]. Available from: http://diabetes.niddk.nih.gov/
11. Nwasuruba CK, Khan M, Egede LE. Racial/ethnic differences in multiple self-care behaviors in adults with diabetes. J Gen Intern Med
12. Paffenbarger RS, Hyde RT, Wing AL, Hsieh CC. Physical activity, all-cause mortality, and longevity of college alumni. N Engl J Med
13. Rowlands AV, Thomas PWM, Eston RG, Topping R. Validation of the RT3 triaxial accelerometer for the assessment of physical activity. Med Sci Sports Exerc
14. Ryan DH, Espeland MA, Foster GD, et al. Look AHEAD (Action for Health in Diabetes): design and methods for a clinical trial of weight loss for the prevention of cardiovascular disease in type 2 diabetes. Control Clin Trials
15. Samuel-Hodge CD, Keyserling TC, France R, et al. A church-based diabetes self-management education program for African Americans with type 2 diabetes. Prev Chronic Dis
[Internet]. 2006 [cited Jan. 19, 2009]. Available from: http://www.cdc.gov/pcd/issues/2006/jul/05_0085.htm
16. Smith TC, Wingard DL, Smith B, Kritz-Silverstein D, Barrett-Connor E. Walking provides strong protection from cardiovascular disease mortality in older adults with diabetes. J Clin Epidemiol
17. Strath SJ, Bassett DR, Swartz AM. Comparison of MTI accelerometer cut-points for predicting time spent in physical activity. Int J Sports Med
18. Tanasescu M, Leitzmann MF, Rimm EB, Hu FB. Physical activity in relation to cardiovascular disease and total mortality among men with type 2 diabetes. Circulation
19. Thomas DE, Elliott EJ, Naughton GA. Exercise for type 2 diabetes mellitus. Cochrane Database Syst Rev
20. US Department of Health and Human Services. Physical Activity Guidelines for Americans Web site [Internet]. Washington (DC): US Department of Health and Human Services; [cited Jan. 19, 2009]. Available from: http://www.health.gov/paguidelines/summary.aspx
21. US Department of Health and Human Services. Physical Activity Guidelines Advisory Committee Report 2008 Web site [Internet]. Washington (DC): US Department of Health and Human Services; [cited 2009 Jan 19]. Available from: http//www.health.gov/paguidelines/committeereport.aspx
22. Wing RR, Jakicic J, Neiberg R, et al. Fitness, fatness, and cardiovascular risk factors in type 2 diabetes: Look AHEAD Study. Med Sci Sports Exerc