Participants were recruited through flyers posted at public institutions (e.g., medical centers) and by advertisements in the local newspaper. Fifty-one community-dwelling healthy older adults (24 men, 27 women) between the ages of 60 and 72 years provided their written informed consent to participate in this study after experimental procedures were explained. The participants’ baseline characteristics are presented in Table 1. None of the participants had any history of diagnosed neurological (e.g., Parkinson disease) or orthopedic (e.g., joint replacement) disorders that might have affected their ability to perform home-based balance exercises conducted during daily tooth brushing or to perform tests for the assessment of balance and muscle strength. The participants were capable of walking independently without any assistive device and they had no prior experience with the applied tests or exercises. Participants were randomly assigned to an intervention (INT) or a passive control group (CON).
Upon entering the laboratory, participants were asked to fill out a number of questionnaires such as the Mini-Mental State Examination Test (MMSE), the Freiburg questionnaire of physical activity (FQoPA), and the Clock Drawing Test (CDT). The CDT is a sensitive screening test for the evaluation of executive function . The elderly participants were asked to include the numbers of a clock in a given circle to make the circle look like a clock. Thereafter, participants were asked to draw the hands of the clock to a self-selected point in time. The next task was to translate the selected time in digital letters and to document those on the sheet. Depending on the study consulted, inter-rater reliability for the CDT ranges between 75.4 and 99.6 % . Test-retest reliability can be classified as high, with a r-value of 0.90 . Cross-correlation with the MMSE revealed a correlation coefficient of r > .50 . As a result, the test distinguishes between pathological and normal test performance. The MMSE is a valid test of cognitive function. It separates patients with cognitive disturbance from those without such disturbance. Test-retest reliability for the MMSE is high, with r = .89. Cross-correlation with the Wechsler Adult Intelligence Scale score revealed a correlation coefficient of r = .78 . An MMSE total score of < 24 separates patients with dementia or functional psychosis from cognitively independently functioning participants and those with anxiety neurosis or personality disorder. The FQoPA assesses the basic physical activity level (e.g., gardening, climbing stairs), leisure time physical activity (e.g., dancing, bowling), and sport activities (e.g., jogging, swimming) of people aged 18 to 78 years . Age-specific corresponding norm values for total physical activity range between 9.9 and 13.6 h per week . Significant test-retest reliability was reported for the summed physical activity level (r = .56). Data from the scientific literature indicated a significant cross-correlation (r = .42) with maximum oxygen uptake .
Prior to testing, all participants received standardized verbal instructions regarding test procedures with a visual demonstration of the balance and muscle strength tests. Thereafter, participants performed a 5-minutes warm up consisting of bipedal and monopedal balance exercises as well as submaximal stepping and skipping movements. Pre- and post-intervention, tests were conducted for the assessment of static steady-state balance (i.e., Romberg test), dynamic steady-state balance (i.e., 10-m single and dual-task walk test), proactive balance (i.e., TUG, Functional-Reach-Test [FRT]), and muscle strength (i.e., CRT). Balance tests were always conducted in randomized order and prior to the muscle strength test. This test sequence was applied in order to keep the effects of neuromuscular fatigue minimal. All participants received one familiarization trial for each test. Thereafter, one test trial was conducted unless otherwise stated.
Balance exercises while tooth brushing
Participants of the INT-group conducted a lifestyle exercise program which included balance exercises conducted during the daily tooth brushing routine. According to Creeth et al. , the teeth were brushed twice per day (i.e., in the morning, right after getting up and at night, before going to bed) for three minutes each on seven days per week for a duration of eight weeks resulting in a total of 112 exercise sessions. This equals an overall exercise time of 336 min and a weekly exercise time of 42 min. The exercise protocol is illustrated in detail in Table 2. Before the exercise period started, participants of the INT-group received information on how to perform the balance exercises by the authors of this article. Additionally, exercise cards containing pictures and descriptions on how to correctly perform all exercises were provided for the eight weeks exercise period. Moreover, participants of the INT-group were asked during weekly phone calls whether they need more information on how to properly perform the balance exercises. By doing so, we wanted to make sure that the exercises were performed with adequate movement skill competence. All participants had to document the realized exercise sessions in a training log. The balance enhancing exercises were performed barefooted or alternatively with socks under different stance and surface conditions. Progression during the balance exercise program was achieved by continuously reducing the base of support (i.e., from step over tandem to one-legged stance) and by including an unstable surface (i.e., rolled up towel). This exercise sequence for progression in balance training has been validated in a previously published study . The participants of the passive CON-group did not receive any intervention during the study period.
Assessment of static, dynamic, and proactive balance
Test circumstances (e.g., room illumination, temperature, noise) during balance assessment were in accordance with recommendations for posturographic testing . Static steady-state balance was assessed using the Romberg Test . Participants performed four tasks with an increasing level of difficulty: (1) standing in an upright position with feet closed and eyes opened for 10 s without swaying while holding both arms extended in horizontal direction with palms facing upwards; (2) ditto, but with eyes closed; (3) ditto, but eyes opened and feet in tandem stance; (4) ditto, but eyes closed and feet in tandem stance. Standing time during the different test conditions was recorded using a stopwatch to the nearest 0.1 s. Maximal stance time for the fourth task was used for further analysis. Age-specific corresponding norm values are 14.0 to 15.0 s for females and 14.3 to 17.5 s for males . High test-retest reliability has been shown for the Romberg Test (eyes opened, intraclass correlation coefficient [ICC] = 0.86 and eyes closed, ICC = 0.84) and Sharpened Romberg Test (eyes opened, ICC = 0.70 and eyes closed, ICC = 0.91) .
Dynamic steady-state balance was assessed using the 10-m walking test. Participants walked with their own footwear at self-selected speeds, initiating and terminating each walk a minimum of one meter before and after the 10-m walkway to allow sufficient distance to accelerate to and decelerate from a steady state of ambulation across the walkway. Time was recorded with a stopwatch to the nearest 0.1 s. Gait speed (m/s) was determined from the time needed to cover the 10-m walking distance under single-task (walking only) and dual-task conditions (walking while concurrently performing a cognitive interference task or a motor interference task). Age-specific norm values for single-task gait speed are 1.10 m/s for females and 1.12 m/s for males .
Proactive balance was assessed using the TUG and the FRT. The TUG was used as described by Podsiadlo and Richardson . Participants were asked to perform the TUG at their self-selected habitual walking speed. Time was recorded with a stopwatch to the nearest 0.1 s. Participants were seated and instructed to walk three meters, turn around, walk back to the chair and sit down. The stopwatch was started on the command “ready-set-go” and stopped as the participant sat down. Age-specific corresponding norm values are 8.0 to 9.0 s for both sexes . The TUG showed excellent test-retest reliability (ICC = 0.99) in older adults . Proactive balance was further assessed by means of the FRT. The FRT measures the maximal distance one can reach forward beyond arm’s length while maintaining a fixed base of support in the standing position . Maximal reach distance of the right arm was recorded to the nearest 0.5 cm. Age-specific norm values are 29.0 to 30.0 cm for both sexes . The FRT showed excellent test-retest reliability (ICC = 0.92) in older adults . Validity of the FRT has previously been shown by Newton et al.  when testing healthy community-dwelling older adults.
Assessment of muscle strength
The CRT as described by Csuka and McMarty  was used for the assessment of lower limbs muscle strength. More precisely, participants sat on a chair with their arms crossed in front of their chest. On the command ready, set, go, participants stood up and sat down as quickly as possible for five times. Three test trials were performed which were separated by a 1-minute rest interval and the best (least time) out of three trials was used for further analysis. Time was recorded with a stopwatch to the nearest 0.1 s. Age-specific norm values have been reported for females (12.7–13.0 s) and males (8.4–11.6 s) . High test-retest reliability has previously been shown for the CRT (ICC = 0.89) .
Cognitive and motor interference tasks
Dynamic steady-state balance (i.e., 10-m walking test) was also examined while performing a concurrent attention-demanding cognitive or motor interference task. The cognitive interference task comprised an arithmetic task in which the participants loudly recited serial subtractions by three, starting from a randomly selected number between 300 and 900 given by the experimenter . The motor interference task required participants to hold two interlocked sticks steadily in front of their body. One stick was held in each hand, with the elbow in 90-degree flexion. Each stick had a ring at the end with a diameter of four cm, and the rings were interlocked . The participants were advised not to let the rings touch each other. When the dual-task methodology was used, participants were instructed to give equal priority to both tasks in order to create real-life conditions .
Descriptive data are presented as group mean values and standard deviations (SD). After normal distribution was examined and confirmed using the Kolmogorov-Smirnov-Test, an independent samples t-Test was applied to determine significant between group baseline differences. Subsequently, a 2 (groups: INT, CON) × 2 (time: pre, post) analysis of variance (ANOVA) with repeated measures on time was used. The classification of effect sizes was determined by converting partial eta-squared to Cohen’s d. The effect size is a measure of the effectiveness of a treatment and it helps to determine whether a statistically significant difference is a difference of practical concern. According to Cohen , effect sizes can be classified as small (0 ≤ d ≤ 0.49), medium (0.50 ≤ d ≤ 0.79), and large (d ≥ 0.80). The significance level was set at p < .05. An a priori power analysis  with an assumed type I error rate of 0.05 and a type II error rate of 0.20 (80 % statistical power) was conducted for balance measures  and revealed that 25 participants per group would be sufficient for revealing interaction effects. Due to potential drop-outs, a total of 55 older adults were enrolled in this study, 29 in INT and 26 in CON. All analyses were performed using the Statistical Package for Social Sciences (SPSS) version 26.0.