Open Access Study protocol

Accelerometer-based physical activity in a large observational cohort - study protocol and design of the activity and function of the elderly in Ulm (ActiFE Ulm) study

Michael D Denkinger1*, Sebastian Franke2*, Kilian Rapp2, Gudrun Weinmayr2, Enric Duran-Tauleria34, Thorsten Nikolaus1, Richard Peter2 and ActiFE Ulm Study Group1

Author Affiliations

1 Agaplesion Bethesda Clinic, Ulm, Germany

2 Institute of Epidemiology, Ulm University, Ulm, Germany

3 Epidemiology and Public Health Unit, Institut de Prestacions d'Assitència Mèdical Personal Municipal (PAMEM), Barcelona, Spain

4 IMIM - Hospital el Mar, Barcelona, Spain

For all author emails, please log on.

BMC Geriatrics 2010, 10:50  doi:10.1186/1471-2318-10-50

Published: 27 July 2010



A large number of studies have demonstrated a positive effect of increased physical activity (PA) on various health outcomes. In all large geriatric studies, however, PA has only been assessed by interview-based instruments which are all subject to substantial bias. This may represent one reason why associations of PA with geriatric syndromes such as falls show controversial results. The general aim of the Active-Ulm study was to determine the association of accelerometer-based physical activity with different health-related parameters, and to study the influence of this standardized objective measure of physical activity on health- and disability-related parameters in a longitudinal setting.


We have set up an observational cohort study in 1500 community dwelling older persons (65 to 90 years) stratified by age and sex. Addresses have been obtained from the local residents registration offices. The study is carried out jointly with the IMCA - Respiratory Health Survey in the Elderly implemented in the context of the European project IMCA II. The study has a cross-sectional part (1) which focuses on PA and disability and two longitudinal parts (2) and (3). The primary information for part (2) is a prospective 1 year falls calendar including assessment of medication change. Part (3) will be performed about 36 months following baseline. Primary variables of interest include disability, PA, falls and cognitive function. Baseline recruitment has started in March 2009 and will be finished in April 2010.

All participants are visited three times within one week, either at home or in the study center. Assessments included interviews on quality of life, diagnosed diseases, common risk factors as well as novel cognitive tests and established tests of physical functioning. PA is measured using an accelerometer-based sensor device, carried continuously over a one week period and accompanied by a prospective activity diary.


The assessment of PA using a high standard accelerometer-based device is feasible in a large population-based study. The results obtained from cross-sectional and longitudinal analyses will shed light on important associations between PA and various outcomes and may provide information for specific interventions in older people.