Promoting Physical Activity Among Cancer Survivors
Promoting Physical Activity Among Cancer Survivors
The literature identified that only one nationally representative study of U.S. cancer survivors used an objective measure of physical activity and examined its association with biologic markers (Lynch et al., 2010). However, in that study, the only marker investigated was adiposity (i.e., waist circumference and BMI), with the results showing an inverse association between adiposity and physical activity among breast cancer survivors.
The epidemiologic examination reported in the current article will further the work in this area by using an objective measure of physical activity in a national sample of different types of cancer survivors and examining understudied biologic markers (e.g., white blood cells [WBCs], neutrophils, homocysteine).
Data from the 2003–2006 NHANES were used. NHANES is an ongoing study conducted by the Centers for Disease Control and Prevention (CDC) that uses a representative sample of noninstitutionalized U.S. civilians selected by a complex, multistage probability design. Participants were interviewed in their homes and subsequently examined in mobile examination centers. All NHANES study procedures were approved by the National Center for Health Statistics ethics review board, with informed consent obtained from all participants prior to data collection (CDC, 2012).
In the 2003–2006 NHANES cycles, 401 participants reported being diagnosed with a cancer that has been associated with physical activity behavior (i.e., breast, prostate, colon, rectal, lung, endometrium/uterus, ovarian, and pancreatic). Of those, 307 provided data on the covariates (i.e., age, gender, race/ethnicity, BMI, cotinine, poverty-to-income ratio [PIR], and comorbidity index). After excluding those with insufficient accelerometry data (i.e., less than four days of 10 or more hours per day of monitoring), 227 participants remained, with those participants comprising the analytic sample. Participants ranged from age 21–85 years. With regard to the final analytic sample (N = 227) and the 80 participants excluded because of insufficient accelerometry data, no differences existed (p > 0.1) with respect to age, gender, race/ethnicity, BMI, cotinine, PIR, and comorbidity index.
Physical Activity. Participants were asked to wear an ActiGraph 7164 accelerometer on their waist for seven days during all activities, with the exception of water-based activities and while sleeping. The accelerometer measured the frequency, intensity, and duration of physical activity by generating an activity count proportional to the measured acceleration. More details about the mechanics of the ActiGraph 7164 accelerometer can be found in another study (Chen & Bassett, 2005).
Estimates for sedentary behavior were classified as less than or equal to 99 counts per minute; light-intensity physical activity was classified between 100 and 2,019 counts per minute; moderate- to vigorous-intensity physical activity (MVPA) was classified as less than or equal to 2,020 counts per minute. Moderate- and vigorous-intensity physical activity were combined because participants spent little time at vigorous-intensity physical activity (X̄ = 0.39 minutes per day, standard error [SE] = 0.13). All estimates were summarized in one-minute bouts. For the analyses described, and to represent habitual physical activity patterns, only those participants with at least four days of 10 or more hours per day of monitoring data were included in the analyses (Troiano et al., 2008). To determine the amount of time the monitor was worn, nonwear was defined as a minimum of 60 consecutive minutes of zero activity counts, with the allowance of 1–2 minutes of activity counts between 0 and 99 (Troiano et al., 2008).
Biologic and Health Markers. The following biologic and health markers were chosen as they have previously (Humphreys, McLeod, & Ruseski, 2014; Loprinzi & Cardinal, 2012b, 2013; Loprinzi & Pariser, 2013; Loprinzi et al., 2013; Warburton et al., 2010) been associated with physical activity: BMI, waist circumference, systolic and diastolic blood pressure (average of up to four measurements), CRP, HDL cholesterol, fasting low-density lipoprotein (LDL) cholesterol, total cholesterol, fasting triglycerides, fasting glucose, fasting insulin, homocysteine (marker of endothelial function), and blood glycohemoglobin (HbA1C). These measurements were taken in the mobile examination center prior to the measurement of physical activity. Details on the assessment of the variables can be found elsewhere (www.cdc.gov/nchs/nhanes.htm).
Covariates. Information about age, gender, and race/ethnicity were obtained from a questionnaire. As a measure of socioeconomic status, PIR was assessed, with a PIR of less than 1 considered below the poverty threshold. Serum cotinine was measured as a marker of active smoking status or environmental exposure to tobacco (i.e., passive smoking). Serum cotinine was measured by an isotope dilution high-performance liquid chromatography/atmospheric pressure chemical ionization tandem mass spectrometry. BMI was calculated from measured weight and height (weight in kilograms divided by the square of height in meters). A comorbidity index variable was created to classify the number of comorbidities each participant experienced (Charlson, Pompei, Ales, & MacKenzie, 1987; Quan et al., 2011). Participants were classified as having zero, one, two, or three or more comorbidities based on self-report of the following physician-diagnosed chronic diseases or events: arthritis, coronary heart disease, heart attack, congestive heart failure, stroke, emphysema, chronic bronchitis, and hypertension.
All statistical analyses were done using STATA, version 12.0. The analyses accounted for the complex survey design used in NHANES by using survey sample weights, clustering, and primary sampling units. New sample weights were created for the combined NHANES cycles following analytic guidelines for the continuous NHANES. Mobile examination center sample weights were used for all analyses with nonfasting variables; fasting sample weights were used for analyses with fasting variables (i.e., LDL cholesterol, triglycerides, glucose, and insulin).
To examine the association between sedentary behavior and physical activity and the biologic and health variables (which were the dependent variables), multivariate linear regression models were computed. A model was computed for each biologic and health variable. The three accelerometer-determined variables (i.e., sedentary, light, and MVPA) all were included in each model, as well as the covariates of age, gender, race/ethnicity, BMI, cotinine, PIR, and comorbidity index. BMI was not adjusted in the waist circumference model because of collinearity (r = 0.85). In addition, the models with systolic or diastolic blood pressure as the outcome variable controlled for drug therapy (i.e., self-report of taking any hypertensive-lowering medication). The models with HDL cholesterol, LDL cholesterol, triglycerides, or total cholesterol as the outcome variable controlled for drug therapy (i.e., self-report of taking any cholesterol medication). The models with glucose, insulin, or HBA1C as an outcome variable, controlled for drug therapy (i.e., self-report of taking insulin or pills for diabetes).
Among the 227 cancer survivors in the empirical study, the mean age was 68.1 years (SE = 1.1). Of the participants, 69% (SE = 3.7) were female, and 85% (SE = 2.3) were Caucasian. The mean PIR was 2.9 (SE = 0.1), which is well above the poverty threshold (i.e., 1), and the mean BMI was 27.7 (SE = 0.5). The proportion of those with zero, one, two, and three or more comorbidities, respectively, was 20%, 32%, 28%, and 20%. The mean minutes per day of sedentary, light-intensity physical activity, and MVPA, respectively, was 525.3 (SE = 7.8), 295.4 (SE = 6.6), and 11.5 (SE = 1.2). Participants, on average, were cancer survivors for 12.3 years (95% confidence interval [6.0, 18.7]).
Table 1 shows the results of the multivariable linear regression analyses examining the association between the accelerometer-determined variables and the biologic health parameters. Sedentary behavior was not independently associated with any of the biologic parameters. After adjustments, including controlling for sedentary behavior and MVPA, light-intensity physical activity was inversely associated with WBC, neutrophils, insulin, and insulin resistance. After controlling for sedentary behavior, light-intensity physical activity, and other covariates, MVPA was inversely associated with BMI, waist circumference, WBC, and neutrophils, but it was positively associated with HDL cholesterol. Although not a primary objective of this analysis, cotinine was positively associated with neutrophils and inversely associated with HDL cholesterol.
Epidemiologic Examination
The literature identified that only one nationally representative study of U.S. cancer survivors used an objective measure of physical activity and examined its association with biologic markers (Lynch et al., 2010). However, in that study, the only marker investigated was adiposity (i.e., waist circumference and BMI), with the results showing an inverse association between adiposity and physical activity among breast cancer survivors.
The epidemiologic examination reported in the current article will further the work in this area by using an objective measure of physical activity in a national sample of different types of cancer survivors and examining understudied biologic markers (e.g., white blood cells [WBCs], neutrophils, homocysteine).
Methods for the Epidemiologic Examination
Data from the 2003–2006 NHANES were used. NHANES is an ongoing study conducted by the Centers for Disease Control and Prevention (CDC) that uses a representative sample of noninstitutionalized U.S. civilians selected by a complex, multistage probability design. Participants were interviewed in their homes and subsequently examined in mobile examination centers. All NHANES study procedures were approved by the National Center for Health Statistics ethics review board, with informed consent obtained from all participants prior to data collection (CDC, 2012).
In the 2003–2006 NHANES cycles, 401 participants reported being diagnosed with a cancer that has been associated with physical activity behavior (i.e., breast, prostate, colon, rectal, lung, endometrium/uterus, ovarian, and pancreatic). Of those, 307 provided data on the covariates (i.e., age, gender, race/ethnicity, BMI, cotinine, poverty-to-income ratio [PIR], and comorbidity index). After excluding those with insufficient accelerometry data (i.e., less than four days of 10 or more hours per day of monitoring), 227 participants remained, with those participants comprising the analytic sample. Participants ranged from age 21–85 years. With regard to the final analytic sample (N = 227) and the 80 participants excluded because of insufficient accelerometry data, no differences existed (p > 0.1) with respect to age, gender, race/ethnicity, BMI, cotinine, PIR, and comorbidity index.
Measures
Physical Activity. Participants were asked to wear an ActiGraph 7164 accelerometer on their waist for seven days during all activities, with the exception of water-based activities and while sleeping. The accelerometer measured the frequency, intensity, and duration of physical activity by generating an activity count proportional to the measured acceleration. More details about the mechanics of the ActiGraph 7164 accelerometer can be found in another study (Chen & Bassett, 2005).
Estimates for sedentary behavior were classified as less than or equal to 99 counts per minute; light-intensity physical activity was classified between 100 and 2,019 counts per minute; moderate- to vigorous-intensity physical activity (MVPA) was classified as less than or equal to 2,020 counts per minute. Moderate- and vigorous-intensity physical activity were combined because participants spent little time at vigorous-intensity physical activity (X̄ = 0.39 minutes per day, standard error [SE] = 0.13). All estimates were summarized in one-minute bouts. For the analyses described, and to represent habitual physical activity patterns, only those participants with at least four days of 10 or more hours per day of monitoring data were included in the analyses (Troiano et al., 2008). To determine the amount of time the monitor was worn, nonwear was defined as a minimum of 60 consecutive minutes of zero activity counts, with the allowance of 1–2 minutes of activity counts between 0 and 99 (Troiano et al., 2008).
Biologic and Health Markers. The following biologic and health markers were chosen as they have previously (Humphreys, McLeod, & Ruseski, 2014; Loprinzi & Cardinal, 2012b, 2013; Loprinzi & Pariser, 2013; Loprinzi et al., 2013; Warburton et al., 2010) been associated with physical activity: BMI, waist circumference, systolic and diastolic blood pressure (average of up to four measurements), CRP, HDL cholesterol, fasting low-density lipoprotein (LDL) cholesterol, total cholesterol, fasting triglycerides, fasting glucose, fasting insulin, homocysteine (marker of endothelial function), and blood glycohemoglobin (HbA1C). These measurements were taken in the mobile examination center prior to the measurement of physical activity. Details on the assessment of the variables can be found elsewhere (www.cdc.gov/nchs/nhanes.htm).
Covariates. Information about age, gender, and race/ethnicity were obtained from a questionnaire. As a measure of socioeconomic status, PIR was assessed, with a PIR of less than 1 considered below the poverty threshold. Serum cotinine was measured as a marker of active smoking status or environmental exposure to tobacco (i.e., passive smoking). Serum cotinine was measured by an isotope dilution high-performance liquid chromatography/atmospheric pressure chemical ionization tandem mass spectrometry. BMI was calculated from measured weight and height (weight in kilograms divided by the square of height in meters). A comorbidity index variable was created to classify the number of comorbidities each participant experienced (Charlson, Pompei, Ales, & MacKenzie, 1987; Quan et al., 2011). Participants were classified as having zero, one, two, or three or more comorbidities based on self-report of the following physician-diagnosed chronic diseases or events: arthritis, coronary heart disease, heart attack, congestive heart failure, stroke, emphysema, chronic bronchitis, and hypertension.
Data Analysis
All statistical analyses were done using STATA, version 12.0. The analyses accounted for the complex survey design used in NHANES by using survey sample weights, clustering, and primary sampling units. New sample weights were created for the combined NHANES cycles following analytic guidelines for the continuous NHANES. Mobile examination center sample weights were used for all analyses with nonfasting variables; fasting sample weights were used for analyses with fasting variables (i.e., LDL cholesterol, triglycerides, glucose, and insulin).
To examine the association between sedentary behavior and physical activity and the biologic and health variables (which were the dependent variables), multivariate linear regression models were computed. A model was computed for each biologic and health variable. The three accelerometer-determined variables (i.e., sedentary, light, and MVPA) all were included in each model, as well as the covariates of age, gender, race/ethnicity, BMI, cotinine, PIR, and comorbidity index. BMI was not adjusted in the waist circumference model because of collinearity (r = 0.85). In addition, the models with systolic or diastolic blood pressure as the outcome variable controlled for drug therapy (i.e., self-report of taking any hypertensive-lowering medication). The models with HDL cholesterol, LDL cholesterol, triglycerides, or total cholesterol as the outcome variable controlled for drug therapy (i.e., self-report of taking any cholesterol medication). The models with glucose, insulin, or HBA1C as an outcome variable, controlled for drug therapy (i.e., self-report of taking insulin or pills for diabetes).
Results of the Epidemiologic Examination
Among the 227 cancer survivors in the empirical study, the mean age was 68.1 years (SE = 1.1). Of the participants, 69% (SE = 3.7) were female, and 85% (SE = 2.3) were Caucasian. The mean PIR was 2.9 (SE = 0.1), which is well above the poverty threshold (i.e., 1), and the mean BMI was 27.7 (SE = 0.5). The proportion of those with zero, one, two, and three or more comorbidities, respectively, was 20%, 32%, 28%, and 20%. The mean minutes per day of sedentary, light-intensity physical activity, and MVPA, respectively, was 525.3 (SE = 7.8), 295.4 (SE = 6.6), and 11.5 (SE = 1.2). Participants, on average, were cancer survivors for 12.3 years (95% confidence interval [6.0, 18.7]).
Table 1 shows the results of the multivariable linear regression analyses examining the association between the accelerometer-determined variables and the biologic health parameters. Sedentary behavior was not independently associated with any of the biologic parameters. After adjustments, including controlling for sedentary behavior and MVPA, light-intensity physical activity was inversely associated with WBC, neutrophils, insulin, and insulin resistance. After controlling for sedentary behavior, light-intensity physical activity, and other covariates, MVPA was inversely associated with BMI, waist circumference, WBC, and neutrophils, but it was positively associated with HDL cholesterol. Although not a primary objective of this analysis, cotinine was positively associated with neutrophils and inversely associated with HDL cholesterol.