Short Spurts of Physical Activity and Cardiovascular Health
Short Spurts of Physical Activity and Cardiovascular Health
We used data from the Coronary Artery Risk Development in Young Adults (CARDIA) study, which has been described in detail elsewhere. In brief, CARDIA is a population-based prospective epidemiologic study of the predictors and development of CVD risk factors and subclinical CVD in young adults. Study participants were 18–35 yr of age at the baseline examination that occurred from 1985 to 1986. For the current analysis using the 20-yr CARDIA follow-up visit, the age range was 37–55 yr of age. Subjects were enroled in approximately equal proportions of race (White/Black), sex, education (high school or lower or higher than high school), and age (≤24 and >24 yr) from Birmingham, AL, Chicago, IL, Minneapolis, MN, and Oakland, CA. All CARDIA study participants provided an informed consent, and each study site's institutional review board approved the parent study.
This particular analysis focused on a subset of the CARDIA cohort that attended both the year 20 (2005–2006) and year 25 (2010–2011) examinations. The year 20 and 25 examinations were attended by 3549 and 3498 study participants, respectively. The year 20 examination was the first study visit in which objectively measured physical activity was collected. Of the 3549 study participants who attended the year 20 examination, 41% or 2076 had valid accelerometer data and comprised our analytic cohort. For this secondary data analysis, the 20-yr examination was considered our study "baseline" and the 25-yr examination was the "5-yr follow-up."
Physical activity was measured with an ActiGraph monitor (model 7164; Pensacola, FL), a uniaxial accelerometer. CARDIA study participants were asked to wear an accelerometer around their waist except when sleeping or bathing for seven consecutive days after the 20-yr clinic visit. Data were collected and expressed as 60-s epochs.
Data from the accelerometer were downloaded and screened for wear time using methods reported by Troiano et al.. Briefly, device nonwear was defined as 60 consecutive minutes of 0 counts, with an allowance for 1–2 min of detected counts between 0 and 100. Wear time was determined by subtracting derived nonwear time from 24 h. A minimum of 10 h·d of wear time on at least 4 of 7 d was necessary to be included in analyses.
MVPA was defined by the Freedson criteria, i.e., activity counts ≥1952. Minutes of MVPA was classified into either 1) bouted MVPA, i.e., activity lasting ≥10 continuous minutes or 2) short spurts of MVPA, i.e., activity lasting <10 continuous minutes. Bouts were defined in accordance with the National Health and Nutrition Examination Survey criteria for a modified 10-min bout. In particular, bouted MVPA was defined as ≥10 consecutive minutes above the Freedson criteria, with allowance for 1- or 2-min interruptions below the 1952 count threshold. Time spent in short spurts was defined as MVPA not meeting the modified 10-min bout criteria (short spurts of MVPA = total time in MVPA - time spent in bouted MVPA). Minutes in bouted MVPA and short spurts of MVPA were averaged over the number of valid days and expressed as minutes per day. For analysis, we classified data in three ways, as follows: 1) continuously, into 10-min increments of time spent in short spurts of MVPA and bouted MVPA, 2) categorically, into separate tertiles of time spent in short spurts of MVPA and bouted MVPA, and 3) a nine-level categorical variable using each pair of bouted and short spurts of MVPA tertiles. The reference value for the nine-level categorical variable was the combination of the tertile for the least time spent in bouted MVPA and the tertile for the least time spent in short spurts of MVPA.
Blood pressure was measured using a standardized protocol. Briefly, blood pressure was measured three times at 1-min intervals after study participants sat for 5 min. The average of the second and third blood pressure measures was used for analysis. Incident hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure, ≥90 mm Hg, or initiation of treatment with antihypertensive medications at follow-up among study participants not meeting these criteria at baseline.
Height and body weight were measured using a standardized protocol, with participants wearing light clothing without shoes. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared. Incident obesity was defined as BMI ≥ 30 kg·m at follow-up among those with BMI < 30 kg·m at baseline.
The following factors were considered as potential confounders (on the basis of their association with physical activity and cardiovascular risk factors in previous studies: age, sex, BMI, race (Black or White), and income (<$16,000, $16,000–$50,000, and ≥$50,000). Age, sex, race, and income data were obtained from standardized self-report questionnaires.
We computed means, SD, medians, and 25th and 75th percentile values for continuous variables and proportions for categorical variables to describe the study sample. We also evaluated the correlation between bouted MVPA and short spurts of MVPA as continuous measures using Pearson correlation coefficient after confirming that data were normally distributed.
We examined the association of short spurts of MVPA with the incidence of obesity and hypertension determined at the 5-yr follow-up by calculating risk ratios (RR) using binomial regression with robust variance estimation. We conducted these analyses by classifying time in short spurts of MVPA as tertiles and as a continuous measure, i.e.,10-min increments, in separate models. All models were adjusted for bouted MVPA and potential confounders. For incident obesity, we did not adjust for baseline BMI because the bias introduced by such an adjustment may exceed the bias eliminated.
To disentangle confounding due to potential correlation between bouted MVPA and short spurts of MVPA tertiles, we evaluated the combined association of MVPA types with the incidence of hypertension and obesity. We calculated the association of each pair of bouted and short spurts of MVPA tertiles with our study outcomes. The tertile pair of the highest values of both bouted and short spurts of MVPA was the reference group. Similar approaches using combined associations with correlated measures have been applied with cardiorespiratory fitness and body fat change, sitting and physical activity, and gait speed and walking endurance.
Methods
We used data from the Coronary Artery Risk Development in Young Adults (CARDIA) study, which has been described in detail elsewhere. In brief, CARDIA is a population-based prospective epidemiologic study of the predictors and development of CVD risk factors and subclinical CVD in young adults. Study participants were 18–35 yr of age at the baseline examination that occurred from 1985 to 1986. For the current analysis using the 20-yr CARDIA follow-up visit, the age range was 37–55 yr of age. Subjects were enroled in approximately equal proportions of race (White/Black), sex, education (high school or lower or higher than high school), and age (≤24 and >24 yr) from Birmingham, AL, Chicago, IL, Minneapolis, MN, and Oakland, CA. All CARDIA study participants provided an informed consent, and each study site's institutional review board approved the parent study.
This particular analysis focused on a subset of the CARDIA cohort that attended both the year 20 (2005–2006) and year 25 (2010–2011) examinations. The year 20 and 25 examinations were attended by 3549 and 3498 study participants, respectively. The year 20 examination was the first study visit in which objectively measured physical activity was collected. Of the 3549 study participants who attended the year 20 examination, 41% or 2076 had valid accelerometer data and comprised our analytic cohort. For this secondary data analysis, the 20-yr examination was considered our study "baseline" and the 25-yr examination was the "5-yr follow-up."
Physical Activity
Physical activity was measured with an ActiGraph monitor (model 7164; Pensacola, FL), a uniaxial accelerometer. CARDIA study participants were asked to wear an accelerometer around their waist except when sleeping or bathing for seven consecutive days after the 20-yr clinic visit. Data were collected and expressed as 60-s epochs.
Data from the accelerometer were downloaded and screened for wear time using methods reported by Troiano et al.. Briefly, device nonwear was defined as 60 consecutive minutes of 0 counts, with an allowance for 1–2 min of detected counts between 0 and 100. Wear time was determined by subtracting derived nonwear time from 24 h. A minimum of 10 h·d of wear time on at least 4 of 7 d was necessary to be included in analyses.
MVPA was defined by the Freedson criteria, i.e., activity counts ≥1952. Minutes of MVPA was classified into either 1) bouted MVPA, i.e., activity lasting ≥10 continuous minutes or 2) short spurts of MVPA, i.e., activity lasting <10 continuous minutes. Bouts were defined in accordance with the National Health and Nutrition Examination Survey criteria for a modified 10-min bout. In particular, bouted MVPA was defined as ≥10 consecutive minutes above the Freedson criteria, with allowance for 1- or 2-min interruptions below the 1952 count threshold. Time spent in short spurts was defined as MVPA not meeting the modified 10-min bout criteria (short spurts of MVPA = total time in MVPA - time spent in bouted MVPA). Minutes in bouted MVPA and short spurts of MVPA were averaged over the number of valid days and expressed as minutes per day. For analysis, we classified data in three ways, as follows: 1) continuously, into 10-min increments of time spent in short spurts of MVPA and bouted MVPA, 2) categorically, into separate tertiles of time spent in short spurts of MVPA and bouted MVPA, and 3) a nine-level categorical variable using each pair of bouted and short spurts of MVPA tertiles. The reference value for the nine-level categorical variable was the combination of the tertile for the least time spent in bouted MVPA and the tertile for the least time spent in short spurts of MVPA.
Hypertension
Blood pressure was measured using a standardized protocol. Briefly, blood pressure was measured three times at 1-min intervals after study participants sat for 5 min. The average of the second and third blood pressure measures was used for analysis. Incident hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure, ≥90 mm Hg, or initiation of treatment with antihypertensive medications at follow-up among study participants not meeting these criteria at baseline.
Obesity
Height and body weight were measured using a standardized protocol, with participants wearing light clothing without shoes. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared. Incident obesity was defined as BMI ≥ 30 kg·m at follow-up among those with BMI < 30 kg·m at baseline.
Potential Confounders
The following factors were considered as potential confounders (on the basis of their association with physical activity and cardiovascular risk factors in previous studies: age, sex, BMI, race (Black or White), and income (<$16,000, $16,000–$50,000, and ≥$50,000). Age, sex, race, and income data were obtained from standardized self-report questionnaires.
Statistical Analysis
We computed means, SD, medians, and 25th and 75th percentile values for continuous variables and proportions for categorical variables to describe the study sample. We also evaluated the correlation between bouted MVPA and short spurts of MVPA as continuous measures using Pearson correlation coefficient after confirming that data were normally distributed.
We examined the association of short spurts of MVPA with the incidence of obesity and hypertension determined at the 5-yr follow-up by calculating risk ratios (RR) using binomial regression with robust variance estimation. We conducted these analyses by classifying time in short spurts of MVPA as tertiles and as a continuous measure, i.e.,10-min increments, in separate models. All models were adjusted for bouted MVPA and potential confounders. For incident obesity, we did not adjust for baseline BMI because the bias introduced by such an adjustment may exceed the bias eliminated.
To disentangle confounding due to potential correlation between bouted MVPA and short spurts of MVPA tertiles, we evaluated the combined association of MVPA types with the incidence of hypertension and obesity. We calculated the association of each pair of bouted and short spurts of MVPA tertiles with our study outcomes. The tertile pair of the highest values of both bouted and short spurts of MVPA was the reference group. Similar approaches using combined associations with correlated measures have been applied with cardiorespiratory fitness and body fat change, sitting and physical activity, and gait speed and walking endurance.