completion of the study to document their usual daily activity level. Patients in the intervention
group trained their lower body on a movement trainer (ReckMOTOmed) with a computer
controlled and individually preassigned training flow. Caregivers were asked to choose a familiar
chair prior to study begin. The movement trainer was positioned in front of the patient and
PA training was performed from the comfort of that chair. We anticipated that using the familiar
chair would encourage the patient’s participation and reassure the caregiver that PA would
not cause adverse reactions, e.g. falls. Participants were required to train three times a week for
30 minutes at an individually chosen time with at least one day without training in between
two training days. Times and dates of PA training were documented by the movement trainer
computer system. Patients were considered to have successfully completed the intervention if
they trained for at least 75% of the time required by the protocol. The program changed between
passive, motor-assisted or active resistive training of the legs as well as changes in direction
(forward, reverse) every 5 minutes. During a familiarization session prior to the study
patients were asked to use the movement trainer and to respond to whether they felt that the
training resistance level was appropriate (too easy, too high, just right). They started with level
1 and the level was increased as they made their choice. The movement trainer has 20 levels of
motor resistance and the patients in the intervention group chose an activity level between 2
and 4. The level did not change over time. Caregivers were asked to act encouraging but to
leave the room as soon as the patient started training. This design was chosen to control for social
contact times between both groups. The movement trainer was removed from the patients’
homes after completion of the 12-week PA program. The control group received the same
monthly clinical visits and a counselling by the treating physician, which included specific advice
how to change inactive habits and increase the PA level. All participants underwent testing
by a blinded psychologist at baseline, 12 and 24 weeks (at 10 a.m.). Measures included activities
of daily living (ADL, ADCS ADL total score) and behavioural symptoms of dementia (NPI
total score) and caregiver burden (NPI total burden score). Cognitive evaluation included the
MMSE and measures of executive function and language ability applying the semantic and
phonemic word fluency as measured by the CERAD [17] and the FAS-test [18]. The data was
compared to age and education corrected normative data. Reaction time, hand-eye quickness
and attention were measured using the reaction time ruler or FETZ-test [19]. The test determines
how long it takes for a patient to respond to the dropping of a ruler with a length of
0.50m. The patient is asked to hold the ruler with his thumb and forefinger and to release the
ruler while the investigator continues to hold it. The patient is instructed to catch the ruler as
fast as he can, as soon as the investigator releases it. The number displayed on the ruler right
over the patient’s thumb is noted. During the visits caregivers were interviewed and asked to report
their experience with the training program and with leaving the patients alone during PA
(intervention group) or their progress in daily PA (controls).