For risk of falling, the total sample size by pooling 8 trials was 1579. For rate of falls, the total sample size by pooling 7 trials was 2012. Weights are from random-effects analysis. Box sizes correspond to precision; the bigger the box, the more precise. Precision was determined as the inverse of the variance. IRR indicates incidence rate ratio; RR, risk ratio.
Effect sizes are Hedges g standardized mean differences (SMDs). Weights are from random-effects analysis. Box sizes correspond to precision; the bigger the box, the more precise. Precision was determined as the inverse of the variance for each estimate. For balance (A), assessments included the Berg Balance Scale, 1-leg stance, and functional reach. For the Berg Balance Scale and 1-leg stance, an increase of time indicates an improvement in the test result. For functional reach, an increase in reached distance indicates an improvement. Hence, positive values would favor the intervention groups over the control groups. For mobility (B), all included trials assessed mobility with the Timed Up and Go test. Decreased completion time indicated improvement in the Time Up and Go Test result. Hence, negative values would favor the intervention groups over the control groups.
Effect sizes are Hedges g standardized mean differences (SMDs). Weights are from random-effects analysis. Box sizes correspond to precision; the bigger the box, the more precise. Precision was determined as the inverse of the variance for each estimate. For lower body strength (A), assessments included the 5 times sit to stand (STS) test (the time to complete 5 stands is measured) and the 30-second STS test (number of stands completed within 30 seconds is measured). Effect sizes of the 5 times STS test were multiplied by −1 to achieve same direction signaling improvement as the effect sizes of the 30-second STS test. Hence, positive values would favor the intervention groups over the control groups. For upper body strength (B), all included trials assessed UBS using handgrip strength. Increasing values indicate an improvement in the test result. Hence, positive values would favor the intervention groups over the control groups.
eAppendix. Supplementary Material
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Mattle M, Chocano-Bedoya PO, Fischbacher M, et al. Association of Dance-Based Mind-Motor Activities With Falls and Physical Function Among Healthy Older Adults: A Systematic Review and Meta-analysis. JAMA Netw Open. 2020;3(9):e2017688. doi:10.1001/jamanetworkopen.2020.17688
Are dance-based mind-motor activities associated with benefits for fall prevention and better physical functions, such as balance, mobility, and strength, in healthy adults 65 years and older?
This systematic review and meta-analysis of 29 randomized clinical trials found that dance-based mind-motor interventions were associated with a statistically significant reduction (37%) in fall risk and a statistically significant reduced rate (31%) of falls. There was a statistically significant association between favorable physical function outcomes and dance-based mind-motor activities for participants in the dance-based mind-motor intervention groups compared with those in the control groups.
Dance-based mind-motor activities may help fall prevention efforts in healthy older adults.
Falls increase morbidity and mortality in adults 65 years and older. The role of dance-based mind-motor activities in preventing falls among healthy older adults is not well established.
To assess the effectiveness of dance-based mind-motor activities in preventing falls.
Systematic search included the PubMed, Embase, Cochrane Library, Web of Science, CINAHL, PsychINFO, Abstracts in Social Gerontology, AgeLine, AMED, and Scopus databases from database inception to February 18, 2018, using the Medical Subject Headings aged 65 and older, accidental falls, and dancing.
This systematic review and meta-analysis included 29 randomized clinical trials that evaluated a dance-based mind-motor activity in healthy older adults with regard to fall risk, fall rate, or well-established measures of physical function in the domains of balance, mobility, and strength. The included studies targeted participants without comorbidities associated with higher fall risk. Dance-based mind-motor activities were defined as coordinated upright mind-motor movements that emphasize dynamic balance, structured through music or an inner rhythm (eg, breathing) and distinctive instructions or choreography, and that involve social interaction.
Data Extraction and Synthesis
Standardized independent screening, data extraction, and bias assessment were performed. Data were pooled using random-effects models. The study followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guideline.
Main Outcomes and Measures
Primary outcomes were risk of falling and rate of falls. For the secondary end points of physical function (balance, mobility, and strength), standardized mean differences (SMDs) were estimated and pooled (Hedges g).
In this systematic review and meta-analysis of 29 randomized clinical trials, dance-based mind-motor activities were significantly associated with reduced (37%) risk of falling (risk ratio, 0.63; 95% CI, 0.49-0.80; 8 trials, 1579 participants) and a significantly reduced (31%) rate of falls (incidence rate ratio, 0.69; 95% CI, 0.53-0.89; 7 trials, 2012 participants). In addition, dance-based mind-motor activities were significantly associated with improved physical function in the domains of balance (standardized mean difference [SMD], 0.62; 95% CI, 0.33-0.90; 15 trials, 1476 participants), mobility (SMD, −0.56; 95% CI, −0.81 to −0.31; 13 trials, 1379 participants), and lower body strength (SMD, 0.57; 95% CI, 0.23-0.91; 13 trials, 1613 participants) but not upper body strength (SMD, 0.18; 95% CI, −0.03 to 0.38; 4 trials, 414 participants).
Conclusion and Relevance
Among healthy older adults, dance-based mind-motor activities were associated with decreased risk of falling and rate of falls and improved balance, mobility, and lower body strength. This type of activity may be useful in preventing falls in this population.
In industrialized countries, life expectancy and the number of age-related chronic diseases are increasing.1 Currently, approximately 30% of community-dwelling adults 65 years and older experience a fall per year, increasing up to approximately 50% of adults 80 years and older.2-4 Notably, more than 30% of falls among older adults need medical attention, and approximately 5% to 7% of falls result in a fracture.4,5 Consequently, falls are costly6 and carry a high risk of functional decline and loss of autonomy for the individual.4
On the other hand, exercise interventions are effective in improving strength, balance, and gait and reducing falls and related injuries among older adults.7,8 Dance-based mind-motor activities have been suggested as physical exercise9-12 with extended benefits beyond the physical on cognition,13,14 social interaction,15,16 quality of life,17 and motivation to be physically active.18-20 Mechanistically, these benefits may be explained by the multidimensional nature of these activities, which combine sensorimotor and cognitive engagement,21,22 relevant to fall prevention.
To our knowledge, no prior meta-analysis of randomized clinical trials (RCTs) has reported on associations between dance-based mind-motor activities and the risk of falling, the rate of falls, and physical function outcomes (balance, mobility, and strength). Therefore, we conducted a systematic review and meta-analysis to summarize the current evidence from RCTs among healthy adults 65 years and older.
For this systematic review and meta-analysis, we conducted a systematic search in PubMed, Embase, the Cochrane Library, Web of Science, CINAHL, PsychINFO, Abstracts in Social Gerontology, AgeLine, AMED, and Scopus databases from database inception to February 18, 2018, using thesaurus terms and key words (eAppendix in the Supplement for search terms used in PubMed). In addition, we performed lateral screening of the bibliographies of reviewed publications. We contacted authors of articles without sufficient information for complete data.
Screening and data extraction were predefined and standardized and followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline.23 At every stage, 2 independent researchers (from among M.M., M.F., U.M., and R.M.) screened each article. Disagreements were solved by consensus and, if necessary, by the consultation of a third independent reviewer (P.O.B.-C.). We included only RCTs that analyzed the effect of dance-based mind-motor activities on the number of persons who fell (risk of falling) or on the number of falls (rate of falls) as primary outcomes or at least 1 of the selected secondary outcomes related to physical function (balance, mobility, or strength) compared with an active (exercise not related to dance-based mind-motor activities) or a passive control group.
We defined dance-based mind-motor activities as coordinated upright mind-motor movements that emphasize dynamic balance, structured through music or an inner rhythm (eg, breathing) and distinctive instructions or choreography, and involve social interaction. Mind-motor activities are activities that combine cognitive and physical tasks (dual tasking or multitasking activities) that involve working memory and deliberate motor control. Dance-based mind-motor activities can be performed solo, in a pair, or in group formations. In addition to several styles of well-known dance-based mind-motor activities, such as folk or ballroom dancing, tai chi fulfills the above definition. Contrarily, most yoga styles focus on static body postures rather than dynamic movements and therefore were excluded. Furthermore, we excluded stepping exercises on so-called dance platforms because of the lack of social interaction.
We only considered RCTs that included healthy and ambulatory older adults (mean age, ≥65 years) living in the community or in independent living facilities. We excluded studies involving participants with comorbidities that directly affect the risk of falling (eg, Parkinson disease, type 2 diabetes, and cognitive decline). Only articles in the English language fulfilled the eligibility criteria. Two reviewers (from among M.M., M.F., U.M., and P.O.B.-C.) independently assessed potential sources of bias using the Cochrane Collaboration’s tool for assessing risk of bias.24 All included studies had been approved by an ethics committee and informed consent of participants had been obtained.
To determine the strength of associations between a dance-based mind-motor activity and risk of falling and rate of falls, we pooled risk ratios (RRs) for risk of falling and Poisson-derived incidence rate ratios (IRRs) for the rate of falls using random-effects models based on intention-to-treat analysis. We estimated the IRRs based on reported incident falls and person-years in trials that did not directly report IRRs.
For the secondary outcomes, only trials that reported standardized, validated, and clinically used tests to measure balance, mobility, or strength were included: the Berg Balance Scale,25 the 1-leg test,26 and the functional reach test27 for balance; Timed Up and Go test28 for mobility; and sit to stand tests29,30 for lower body strength and the handgrip strength test31 for upper body strength. To determine the strength of associations between a dance-based mind-motor activity and physical function measures, we calculated a pooled Hedges g standardized mean difference (SMD) using a random-effects model.32 If a study reported outcomes stratified or by 2 tests that assessed the same outcome, we pooled the reported effect sizes before entering them into the meta-analysis. We weighted stratified outcomes by the inverse of the variance, assumed dependency between the assessments, and used a conservative correlation coefficient of 0.8.32 To interpret the effect of the intervention, we followed the guideline suggested by Cohen: equivalent effect size (SMD of 0), small effect size (SMD of 0.2), medium effect size (SMD of 0.5), and large effect size (SMD of 0.8).33
Heterogeneity was quantified by the I2 statistic as suggested by Borenstein et al.32 We performed prespecified subgroup analyses by type of dance-based mind-motor intervention (non–tai chi vs tai chi), frequency (<3 vs ≥3 times per week), duration (<12 vs ≥12 weeks), type of dwelling (community dwelling vs independent living facilities), and type of randomization (participants randomized vs cluster randomization). In addition, we performed a random-effects meta-regression by intervention frequency and intervention duration if more than 10 RCTs were included.
Small-study effects were assessed for each outcome if more than 10 RCTs were available. We conducted the Harbord modified test for the primary outcome of RR of falls34 and the Egger test for balance, mobility, and lower body strength.32 We used the Egger publication bias plot and funnel plots to visually assess indication of publication bias.35 We used Stata software, version 15 (StataCorp LLC) for data analysis.36 Statistical assessment was 2 sided and considered statistically significant at P < .05.
From 4627 screened publications, 29 RCTs that fulfilled the inclusion criteria were identified (eAppendix in the Supplement). Eight trials37-44 reported on the risk of falling, 7 trials40-43,45-47 reported on the rate of falls, and 4 trials40-43 reported on both. A total of 15 trials37-41,44,48-56 reported on balance, 13 trials39-41,45,48,50,54,57-62 reported on mobility, 13 trials38,45,46,50,56,58-65 reported on lower body strength, and 4 trials44,47,55,59 reported on upper body strength. Table 1 gives an overview of all included trials. Table 2 lists the main descriptive statistics by outcome.
Overall, the sample size of the trials varied greatly, from 19 up to 684 participants. Most trials38-42,44,45,47-65 included community-dwelling older adults, with most participants being women. Seven trials37,40,41,43,47,52,64 reported the mean age of their population being at least 75 years. Three trials37,43,46 were conducted among participants living in independent living facilities. Five trials37,39,43,46,60 used cluster randomization.
Within the 29 trials, the interventions lasted between 6 weeks and 12 months, and the duration of intervention sessions ranged from 35 to 120 minutes. The frequency of the study intervention varied between once per week41,42 and 4 times per week,51 with most trials reporting 2 times per week43,46-49,52,55,56,60-62,64,65 or 3 times per week.37-40,44,45,50,53,54,57-59,63 Adherence was at least 80% in 15 of the 22 trials40,41,44,45,48,49,51,52,54,57-61,64 that reported adherence.
In terms of type of dance-based mind-motor activity, 13 trials38,41,46,48-50,54,57,58,60,63-65 investigated activities that involved dance styles, such as ballroom or folk dances, and 16 trials37,39,40,42-45,47,51-53,55,56,59,61,62 investigated tai chi.
More than half of the trials were conducted in North America40,43,45,47-50,56,59,64 (10 trials) or Asia37,39,44,51,55,60,63 (7 trials). Five trials each were from Europe38,41,52,54,58 and Oceania,42,46,61,62,65 and 2 trials53,57 were from South America.
Dance-based mind-motor activities were associated with 37% reduction in risk of falling (RR, 0.63; 95% CI, 0.49-0.80) (Figure 1A) based on 8 trials37-44 of 1579 participants, with a weighted mean (SD) age of 73.2 (4.9) years, a median sample size of 127 (range, 30-684), and median intervention duration of 22 weeks (range, 8-48 weeks).
Prespecified subgroup analyses revealed a significant association with reduction of risk of falling for trials that conducted interventions 3 times per week or more (RR, 0.47; 95% CI, 0.31-0.72) and for trials with durations between 12 and 24 weeks (RR, 0.71; 95% CI, 0.58-0.86 (eAppendix in the Supplement).
Two sensitivity analyses were performed for risk of falling, one excluding the trial by Woo et al,44 which was driving the pooled result, and one excluding the trial by Eyigor et al,38 from which unpublished results were taken. For both sensitivity analyses, the pooled RR remained statistically significant in favor of the intervention groups (RR, 0.74; 95% CI, 0.64-0.86 for the analysis excluding the trial by Woo et al,44 and RR, 0.64; 95% CI, 0.50-0.82 for the analysis excluding the trial by Eyigor et al38) (eAppendix in the Supplement).
For rate of falls, dance-based mind-motor activities were associated with a reduction by 31% (IRR, 0.69; 95% CI, 0.53-0.89; 7 trials) (Figure 1B) based on 7 trials40-43,45-47 of 2012 participants with a weighted mean (SD) age of 74.4 (4.3) years, a median sample size of 188 (range, 54-684), and a median intervention duration of 24 weeks (range, 15-52 weeks).
Subgroup analyses suggested this association was most pronounced in trials that conducted the intervention 3 times per week or more (IRR, 0.55; 95% CI, 0.42-0.73) and for trials that lasted between 12 and 24 weeks (IRR, 0.59; 95% CI, 0.49-0.71) (eAppendix in the Supplement).
An association was found between dance-based mind-motor activities and improved balance (SMD, 0.62; 95% CI, 0.33-0.90) (Figure 2A) in 15 trials37-41,44,48-56 of 1476 participants, with a weighted mean (SD) age of 72.3 (4.3) years, a median sample size of 77 (range, 23-269), and a median intervention duration of 16 weeks (range, 8-48 weeks). Subgroup analyses by intervention type found a higher SMD for non–tai chi (SMD, 0.86; 95% CI, 0.33-1.39) activities. Interventions performed 3 times per week or more were associated with a higher SMD (SMD, 0.84; 95% CI, 0.54, 1.14). Sensitivity analyses excluding the trial by Hopkins et al50 (published much earlier than the others) or the trial by Serrano-Guzmán et al54 (driving the result) also found an association between dance-based mind-motor and improved balance (SMD, 0.63; 95% CI, 0.33-0.93 for the analysis excluding the trial by Hopkins et al50 and SMD, 0.51; 95% CI, 0.24-0.78 for the analysis excluding the trial by Serrano-Guzmán et al54).
An association was also found between dance-based mind-motor activities and improved mobility (SMD, –0.56; 95% CI, –0.81 to –0.31) (Figure 2B) based on 13 trials39-41,45,48,50,54,57-62 of 1379 participants, with a weighted mean (SD) age of 73.0 (3.3 years), a median sample size of 54 (range, 19-451), and a median intervention duration of 16 weeks (range, 8-52 weeks). For mobility, decreased completion time indicated improvement in Timed Up and Go test results. Hence, negative values would favor the intervention groups over control groups. Subgroup analyses found positive associations for non–tai chi activities (SMD, –0.79; 95% CI, –1.16 to –0.42) and for activities with 3 sessions or more per week (SMD, –0.76; 95% CI, –1.10 to 0.42). The results remained positively associated within sensitivity analyses that excluded the trial by Hopkins et al50 or the trial driving the results by Cepeda et al57 (SMD, –0.54; 95% CI, –0.80 to –0.28 for the analysis excluding the trial by Hopkins et al50 and SMD, –0.48; 95% CI, –0.71 to –0.25 for the analysis excluding the trial by Cepeda et al57).
Dance-based mind-motor activities were associated with improved lower body strength (SMD, 0.57; 95% CI, 0.23-0.91) (Figure 3A), based on 13 trials38,45,46,50,56,58-65 with 1613 participants with a weighted mean (SD) age of 73.1 (3.3) years, a median sample size of 54 (range, 19-530), and a median intervention duration of 12 weeks (range, 6-52 weeks). Subgroup analyses supported a positive association with improved lower body strength for non–tai chi activities (SMD, 0.86; 95% CI, 0.25-1.46) and intervention frequencies of 3 times per week or more (SMD, 1.04; 95% CI, 0.31-1.77). In a sensitivity analysis excluding the early trial by Hopkins et al,50 the SMD remained significant (SMD, 0.39; 95% CI, 0.12-0.65). No significant association was found between dance-based mind-motor activities and upper body strength (SMD, 0.18; 95% CI, −0.03 to 0.38) (Figure 3B) based on 4 trials44,47,55,59 with 414 participants. For subgroup and sensitivity analyses of physical function outcomes, see the eAppendix in the Supplement.
The limited number of trials did not allow for the assessment of small-study effects for the risk of falling, rate of falls, and upper body strength.24 For balance, visual inspection of the funnel plot suggested asymmetry, but there was no evidence of small-study effects based on the Egger test (intercept, −1.59; SE, 1.45; P = .29). For mobility and lower body strength, the funnel plots showed asymmetry, and the Egger test result was statistically significant (intercept, −2.87; SE, 1.04; P = .02 for mobility and intercept, 4.04; SE, 1.05; P = .003 for lower body strength), suggesting that small-study effects cannot be ruled out.
Moderate heterogeneity was found for the primary outcomes (risk of falling: I2 = 54.8%, P = .03; rate of falls: I2 = 73.0%, P = .001). For secondary outcomes, substantial heterogeneity was found in all domains except upper body strength (balance: I2 = 90.1%, P < .001; mobility: I2 = 77.1%, P < .001; lower body strength: I2 = 88.5%, P < .001; upper body strength: I2 = 0.0%, P = .87).
Based on the Cochrane tool of bias,66 a high risk of bias was found for at least 1 domain in 10 trials37-39,46,49,51,58-60,65 and an unclear risk of bias in at least 1 domain for every included trial (eAppendix in the Supplement).66
In this systematic review and meta-analysis, which included 29 trials of 4239 healthy older adults, there was a significant association between dance-based mind-motor activities, including both tai chi and non–tai chi activities, and reductions in the risk of falling and the rate of falls. The association of dance-based mind-motor activities with consistent improvements in balance, mobility, and lower body strength supports these findings. Notably, the predefined subgroup analyses suggest that greater frequency (≥3 times per week) and a duration of the intervention between 12 and 24 weeks are associated with greater benefits from these interventions with regard to fall and functional outcomes.
For fall prevention, the analysis of dance-based mind-motor interventions are in alignment with findings of prior meta-analyses67,68 of multicomponent physical exercise interventions or tai chi alone. Although the analyses for the primary outcomes included more trials that investigated tai chi than non–tai chi dance-based mind-motor activities, the results by subgroups of type of dance-based mind-motor activities suggest that non–tai chi trials might have similar associations with risk of falls and fall rate reductions. In particular, there was a consistently stronger association with improvement of function and non–tai chi dance-based mind-motor interventions for balance, mobility, and lower body strength, and the results therefore extend and strengthen the evidence base of exercise trials that focus on multitasking skills for fall prevention among healthy older adults.
To our knowledge, this meta-analysis is the first that aims to summarize the associations of dance-based mind-motor activities beyond tai chi with the risk of falling and the rate of falls among healthy older adults. Earlier systematic reviews that investigated dance-based mind-motor activities among healthy older adults suggested a beneficial effect for fall prevention solely based on the improvement of balance or strength9-11,69 or included tai chi interventions only.68 Other meta-analyses on the effect of dance-based mind-motor activities preselected on cardiovascular risk,70 Parkinson disease,71 or cognitive13,72 function. Sherrington et al67 report subgroup analyses only for 1 trial in the category dance, whereas the present study additionally included 2 trials that investigated non–tai chi activities for each outcome (risk of falling38,41 and rate of falls41,46).
Furthermore, the present meta-analysis supports benefits of dance-based mind-motor activities for several dimensions of physical function, including balance, mobility, and lower body strength. The observed nonsignificant association on upper body strength may be explained by the fact that the upper extremities in dance-based mind-motor activities are used for expression and partnering rather than building strength or body weight support.
This study has strengths and limitations. Strengths include the comprehensive search strategy within 10 different databases, which incorporated unpublished information from authors38 and was built on the conceptual framework established by the Prevention of Falls Network Europe (ProFaNE).67 In addition to ProFaNE’s definition of 3-dimensional training, this study combined mind and motor abilities, as well as the domain of social interaction. In this meta-analysis, 13 of 29 trials38,41,46,48-50,54,57,58,60,63-65 tested dance styles.
This study found consistent associations between dance-based mind-motor activities and improvement for falls and physical function, and the associations remained significant for all sensitivity analyses performed with influential studies excluded. Finally, relevant to implementation of dance-based mind-motor activities into public health strategies, most trials included in this meta-analysis reported at least 80% adherence, which may be better than previously reported adherence rates for other structured exercise interventions.20,67
This meta-analysis also has limitations. Because of the limited number of non–tai chi trials, several of these trials had to be excluded, giving a stronger weight to the tai chi interventions. However, subgroup analyses excluding the tai chi trials support a similar positive association of non–tai chi activities with balance, mobility, and lower body strength as tai chi. Although data for risk of falling and fall rate were collected with sufficient quality, aspects of physical function, such as balance and strength, were evaluated in part with nonstandardized and noncomparable assessment devices, such as balance platforms or isokinetic machines, which led to the exclusion of several trials that reported on physical function. Finally, these results might not be generalizable to older men because all but 1 trial included mainly women.39,44,51
Although these results found significant positive associations across risk of falling, rate of falls, and 3 of 4 investigated functional measures and are therefore promising in their consistency and effect size for fall prevention, the study also documented limitations in the quality of individual trials. This is true especially for the non–tai chi interventions. Tai chi is among the best-studied activities for older adults.68 However, additional high-quality trials investigating other types of dance-based mind-motor activities are needed to evaluate options for populations who do not have a tradition of tai chi practice but do have strong cultural bonds to activities such as folk and ballroom dancing or eurhythmics. Additional trials are needed to investigate dance-based mind-motor activities, considering optimal duration and frequency for most effective fall prevention among healthy older adults.
Accepted for Publication: June 29, 2020.
Published: September 25, 2020. doi:10.1001/jamanetworkopen.2020.17688
Correction: This article was corrected on January 22, 2021. An updated supplement has been placed online. In addition, citations in the text, tables, and figures have been corrected, as have several values. The findings of the study were not affected by these changes.
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Mattle M et al. JAMA Network Open.
Corresponding Author: Michèle Mattle, MSc, MPH, Department of Geriatrics and Aging Research, University Hospital Zurich and University of Zurich, c/o Stadtspital Waid, Tièchestrasse 99, 8037 Zürich, Switzerland (firstname.lastname@example.org).
Author Contributions: Ms Mattle had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Mattle, Fischbacher, Meyer, Kressig, Steurer, Bischoff-Ferrari.
Acquisition, analysis, or interpretation of data: Mattle, Chocano-Bedoya, Fischbacher, Meyer, Abderhalden, Lang, Mansky, Kressig, Orav, Bischoff-Ferrari.
Drafting of the manuscript: Mattle.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Mattle, Chocano-Bedoya, Abderhalden, Lang, Orav.
Administrative, technical, or material support: Mattle, Fischbacher.
Supervision: Meyer, Kressig, Steurer, Bischoff-Ferrari.
Conflict of Interest Disclosures: This study was conducted as part of Ms Mattle’s fulfillment of a doctoral degree in clinical sciences at the medical faculty of the University of Zurich, Zurich, Switzerland.
Additional Contributions: Georg Bosshard, MD, Department of Geriatrics and Aging Research, University Hospital Zurich and University of Zurich, Zurich, Switzerland, and Sabine Klein, PhD, University of Zurich, Switzerland, Zurich, Switzerland, supported the writing of the manuscript by providing mentoring, scientific discussion, and systematic database searching. They were not compensated for their work.