Background/Purpose
Compared to the general population, adults with visual impairment display higher incidences of obesity and overweight, co-morbidity, and hypokinetic conditions. Despite these observations, the association between physical activity and health remains understudied in this population. In order to accurately establish the relationship between physical activity and health in adults with visual impairment, a reliable measurement of physical activity must be obtained. Therefore, the purpose of this study was to determine the minimum number of days of monitoring necessary to reliably measure physical activity in adults with visual impairment.
Method
Twenty-seven adults with visual impairment were asked to complete a 7 day step-count assessment using the CYMA Step Activity Monitor; six participants were excluded from further analyses due to inadequate data. For the 21 adults who completed 5 or more days of monitoring, missing days were replaced using an individual-information centered approach by incorporating the average of the remaining values for the participant.
Analysis/Results
The Generalizability theory approach was used to determine the amount of variance in daily step counts associated with the participants in the study (G-study), and to provide a reliability coefficient (G-coefficient) for the minimum timeframe necessary to reliably measure step-based physical activity (D-study). Participant (P) and day (D) were considered random facets in a fully crossed design (P x D) to determine the minimum number of days of data collection necessary to achieve a desirable reliability coefficient (G ≥ .80) for step-count measurement. The G-study illustrated that participants (P) and days (D) accounted for 40% and 3.5% of the total variance, respectively. The interaction between P x D and unidentified error accounted for the majority of the variance (56.5%). The D-study illustrated that a minimum of 6 days were necessary to achieve a G-coefficient ≥ .80.
Conclusions
The between-day variation was found to be minimal; therefore, 6 days of data collection (including week and/or weekend days) may guarantee reliable step-count measurement in adults with visual impairment. Considering that a large proportion of the variance remained unexplained, future investigations in this area should try to identify possible sources of error.