ABSTRACT:
CONTEXT AND OBJECTIVES:
Accelerometry provides objective measurement of physical activity levels, but is unfeasible in clinical practice. Thus, we aimed to identify physical fitness tests capable of predicting physical inactivity among adults.
DESIGN AND SETTING:
Diagnostic test study developed at a university laboratory and a diagnostic clinic.
METHODS:
188 asymptomatic subjects underwent assessment of physical activity levels through accelerometry, ergospirometry on treadmill, body composition from bioelectrical impedance, isokinetic muscle function, postural balance on a force platform and six-minute walk test. We conducted descriptive analysis and multiple logistic regression including age, sex, oxygen uptake, body fat, center of pressure, quadriceps peak torque, distance covered in six-minute walk test and steps/day in the model, as predictors of physical inactivity. We also determined sensitivity (S), specificity (Sp) and area under the curve of the main predictors by means of receiver operating characteristic curves.
RESULTS:
The prevalence of physical inactivity was 14%. The mean number of steps/day (≤ 5357) was the best predictor of physical inactivity (S = 99%; Sp = 82%). The best physical fitness test was a distance in the six-minute walk test and ≤ 96% of predicted values (S = 70%; Sp = 80%). Body fat > 25% was also significant (S = 83%; Sp = 51%). After logistic regression, steps/day and distance in the six-minute walk test remained predictors of physical inactivity.
CONCLUSION:
The six-minute walk test should be included in epidemiological studies as a simple and cheap tool for screening for physical inactivity.
KEY WORDS:
Motor activity; Physical fitness; Accelerometry; ROC curve; Body mass index.
Thumbnail
