An Objective Measure of Physical Activity in the DPPOS: The Accelerometry Ancillary Study Abstract The effectiveness of lifestyle intervention for the prevention of Type2 diabetes has been recently demonstrated in five clinical trials including the Diabetes Prevention Program (DPP). The results of all of these trials support a lifestyle approach (which includes physical activity) to the prevention of Type2 diabetes in individuals that were initially at high risk for diabetes. The uniqueness of the DPP was that it had significant diversity (ethnicity/race, age and geography) among the participants recruited. All of these five clinical trials used a subjective measure to assess physical activity levels in their respective prevention efforts. Physical activity levels in the DPP were shown to be statistically different at the end of the original study between the randomized groups. Partial correlations of activity levels as assessed at baseline by the activity questionnaires were significantly and inversely associated with fasting and 75- g oral glucose tolerance measures of insulin (fasting and 30 minute postload), fasting plasma proinsulin, and HbA1c. Yet, change in physical activity level from baseline to year 2 in the lifestyle arm as determined by theses questionnaires was not significantly related to diabetes incidence after adjustment for weight change. A potential reason for these findings is likely due, in part, to the measurement error of the physical activity questionnaire itself. Modest changes in physical activity levels due to the intervention may not feasibly be detected using subjective measures. Physical activity questionnaires have limitations in assessing low-intensity and unstructured physical activity which have now become part of the focus of current intervention recommendations. Also, as the activity prescription has become more flexible, allowing individuals to split up their activity bouts into smaller pieces, the recall and therefore subjective assessment of physical activity becomes more difficult. Finally, the diversity of the sample also makes accurate assessment across the 27 sites a challenge. For these reasons, in the DPP, to capture the full extent of the physical activity change and its impact, one will likely need to utilize more precise objective measures of physical activity. Activity monitors, more specifically accelerometers, have been successfully used to objectively document physical activity changes in several intervention efforts. The accelerometer is a small activity monitor, typically worn on the hip and is equipped with an electronic sensor that measures both quantity and intensity of movement, resulting in the collection and storage of daily patterns of physical activity. Since physical activity measured by the accelerometer can be partitioned into activity bouts of varying levels of intensity, both the amount of time spent in total movement and the amount of time spent in specific intensities can be determined. We recently analyzed the physical activity data from a national data set (NHANES) in which we demonstrated that two very different conclusions can be drawn in regards to activity ranking of the ethnic groups, depending on which activity assessment is used. Although data from the activity questionnaire identified Hispanic men and women to be among the least active groups, accelerometer data collected around the same time found both Hispanic men and women to be among the most active. In addition, fasting glucose and insulin was shown to be significantly related to accelerometer activity counts but not to total daily activity as assessed by questionnaire. This demonstrates the need to look at activity in the DPP by accelerometer. Since the changes in activity due to the intervention were likely subtle and reflect lifestyle changes hard to determine by questionnaire, and since the DPP population was very diverse in regards to ethnicity/race, age and geographic location, physical activity levels may be too complex to assess by recall. The purpose of this ancillary study is to objectively assess physical activity levels with the use of an accelerometer in all DPP participants that have chosen to participate in the next cycle of the DPP follow-up study and have not yet converted to diabetes. Even though the impact of the intervention appears to have weakened since the end of the original trial, diabetes incidence and cardiovascular disease and metabolic syndrome risk factors continue to be better in the lifestyle group compare to the other two groups. It is reasonable to hypothesize that differences in physical activity levels among the groups may also exist that could be partially responsible for this better health profile. In addition, after we determine the relationship between accelerometer data and reported physical activity derived from the MAQ activity questionnaire, we would be able to use this information to direct DPP and DPPOS data analyses involving activity levels measured by the MAQ (which has been utilized since baseline of the original DPP clinical trial). This application is contingent on the future funding of the second cycle of the D follow-up.

Public Health Relevance

This application aims to objectively assess physical activity levels with the use of an accelerometer in all DPP participants that have chosen to participate in the next cycle of the DPP follow-up study, and have not yet converted to diabetes. All five clinical trials that utilized a lifestyle intervention approach in the prevention of Type 2 diabetes used a subjective measure to assess physical activity levels in their respective efforts. We will likely never have the opportunity to investigate this question in a randomized clinical trial of diabetes prevention again. Therefore, we have convened a team of investigators to carry out this unique project.

National Institute of Health (NIH)
National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)
Research Project (R01)
Project #
Application #
Study Section
Kidney, Nutrition, Obesity and Diabetes (KNOD)
Program Officer
Hunter, Christine
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Pittsburgh
Public Health & Prev Medicine
Schools of Public Health
United States
Zip Code
Rockette-Wagner, Bonny; Storti, Kristi L; Dabelea, Dana et al. (2017) Activity and Sedentary Time 10 Years After a Successful Lifestyle Intervention: The Diabetes Prevention Program. Am J Prev Med 52:292-299