Physical activity questionnaires specific to other individuals are not currently available.
The aim of this research is to create a valid and reliable activity questionnaire from the prototype items that we developed during Phase I. A multiple regression framework will be used to select and weight items from the prototype questionnaire for the prediction of a criterion activity score. A random sample of 600 individuals age 65 and over will be randomly assigned to one of four study protocols. The protocols differ in the timing of administration of the prototype questionnaire relative to the home visit (physical measure/activity assessment) and by mode of administration of the prototype questionnaire (telephone interview/mail). During the home visit measures of resting heart rate, body mass index blood lipids and sociodemographic information will be obtained. Physical activity will be assessed over 3 consecutive days by objective monitors, activity diaries and a self assessment. Principal components analysis will be used to create a criterion activity score based on the 3 day activity assessments. Multiple regression will be used to select and weight items from the prototype questionnaire to provide maximum predictive accuracy in the criterion measure. Reliability will be assessed by reinterviewing 80 individuals two months after baseline. The validity of the activity scale scores will be evaluated on the basis of correlations with five physiologic measurements. A physical activity questionnaire appropriate for older populations would be of value for survey and epidemiological research as well as physicians, therapists and other s concerned with monitoring the health and functioning of older individuals.
