By Tim Wiedenmann, Steffen Held, Ludwig Rappelt, Martin Grauduszus, Sofie Spickermann, and Lars Donath
Fall prevention is of importance for maintaining independency of daily living, well-being, and quality of life in older adults and given demographic trends, an urgent economic challenge for the healthcare system.
Falls are a serious health concern and a major cause of morbidity and mortality in community-dwelling older adults. About 1 in 3 older adults above the age of 65 fall at least once a year and half of them are recurrent fallers. The prevention of falls is, therefore, not only of importance for maintaining independency of daily living, well-being, and quality of life in older adults but also an economic challenge for the healthcare system. Available clinical practice guidelines for fall prevention in the older population underpin the importance of physical activity and exercise. The majority of the included exercise-based fall prevention studies focus on balance exercises or resistance training. While tendencies favoring balance exercises for the prevention of falls can be observed, it is not entirely clear whether balance exercise alone or in a combination with multiple exercise forms is most effective for reducing fall risk.
Against this background, the aims of this network meta-analysis are: (i) to rank different physical activities based on their effect on fall prevention in older adults and (ii) to analyze which form of exercise is most suitable for fall prevention.
Methods
The studies included in this network meta-analysis were identified through a comprehensive search in 5 biomedical databases (PubMed, SportDiscus, CINAHL, Web of Science and EMBASE). The study authors included randomized controlled trials (RCTs) that compared the occurrence of fall events in older adults who received different interventional treatments.
The following inclusion criteria based on the PICOS approach [population (P), intervention (I), comparators (C), main outcome (O), and study design (S)] were applied: Full-text article published in English in a peer-reviewed journal; participants were community-dwelling, independently living people involved in studies with a mean age of at least 65 years and an age larger than 60 years when subtracting 1 standard deviation from the studies’ mean age, without additional diseases (eg, stroke, chronic stroke, Parkinson’s disease, multiple sclerosis, dementia, hip fractures or other fractures) or an acute or chronic mental or physical illness (such as cancer, depression, mild cognitive impairment, diabetes mellitus, or chronic obstructive pulmonary disease,) (P). All studies that included at least 1 exercise intervention group and 1 control or another exercise intervention group were eligible. Supplement and medication studies were excluded (I). Comparators were groups with no or light physical exercise (C). Documentation of the incidence of falls, to estimate the risk ratio (RR), for at least 6 months, regardless of whether they were documented within the intervention period, as a follow-up after the intervention, or during the intervention and in a follow-up period. A fall was defined as a subject’s unintentionally coming to rest on the ground or at some other lower level, not as a result of a major intrinsic event (eg, stroke or syncope) or overwhelming hazard (O). Furthermore, the studies had to be 2- or multiarmed randomized controlled trails (S). The exclusion criteria were: (1) No adequate control conditions, which made integration into the network impossible, and (2) the use of an alternative supporting structures or systems such as an exoskeleton.
To estimate the effect of exercise on the incidence of fallers, the number of fallers and non-fallers in each intervention group were extracted. For the simplification of the network, similar treatments haven been summarized in (i) Active Control (interventions that are not thought to influence the outcome of falls such as light stretching and relaxation); (ii) Combined Postural Control Training (resistance or endurance training performed additionally to postural control training); (iii) Endurance Training; (iv) Inactive Control; (v) Multifactorial Training (forms of training that included other non-exercise related factors influencing the risk of falls (such as home hazard management and visual, educational or behavioral interventions) in addition to postural control training); (vi) Postural Control Training (balance, coordination and/or multitask training); and (vii) Resistance Training.
The RR was calculated for all interventional treatments by dividing the incidence of the intervention group by the incidence of the reference group or by using RR values if they were provided. Additionally mean error and 95% confidence interval (95%-CI) were evaluated. The estimations of treatment effects were calculated based on a random effects model; the Inactive Control served as the reference treatment. A ranking was created based on the P-score of the individual treatments. A forest plot was created to further visualize the ranking and effects of the treatments. The decomposed Q-statistics (within and between designs) were used to interpret potential heterogeneity and inconsistency. Heterogeneity was further quantified by I2. Funnel plots were created to check potential publication bias.
Results
Seventy-six comparisons from 66 RCTs with 4,420 (61% male) participants aged 77 ± 4 (68–88) years were included in this network meta-analysis. The network model revealed low heterogeneity (I2 = 28.0, 95%CI 1.0 to 47.7%) and inconsistency (Q between designs = 15.1, P = 0.37). Postural control training was found to be most effective in preventing falls (Postural Control Training (home): RR = 0.66, 95%-CI [0.49; 0.88], P-score = 0.97; Postural Control Training: RR = 0.82, 95%-CI [0.75; 0.91], P-score = 0.82). Combined and multifactorial interventions also display a robust but smaller effect (RR = 0.88–0.93, P-score = 0.65–0.47). The ranking of the different treatments is depicted in Figure 1.
Discussion
The key finding was that balance and strength focused exercise modes are the most beneficial for the prevention of fall events.
Among all included exercise modes, balance-type exercises revealed the lowest relative risk for a fall event and thus received the highest P-score, outranking the resistance training and resistance training combined approaches. These findings are aligned with previous findings and reflect the majority of available clinical guidelines but surpass them. This network meta-analysis displays that these reductions in fall risk are also present when postural control training is performed at home and largely unsupervised.
Multifactorial Training, Combined Postural Training (home), and Combined Postural Training are associated with a slight reduction of fall risk and a high precision of data, indicated by narrow confidence limits. These findings suggest that an interventional approach with multiple different exercise modes or other non-exercise related factors are inferior in magnitude but robust in effect occurrence of fall risk reductions compared with the most beneficial postural control training interventions when they are performed isolated. This is at least partially in line with a 2007 meta-regression and meta-analysis in which Campbell et al found that single factorial interventions had similar but slightly favorable effects compared to multifactorial interventions when it comes to the prevention of falls. Although the effects of the combined interventions are inferior to the best single factorial interventions, the high precision of the data combined with the large amount of included evidence (51 direct comparisons) leads to the assumption that some combined and multifactorial interventions might be a valuable alternative. The value of multifactorial interventions might even be higher when the interventions are designed with different domains that are specific to the needs of the individual patient.
In contrast, Resistance Training, Resistance Training (home), Endurance Training, and Multifactorial Training (home) do not have a robust positive effect on the relative risk of falls. While these interventions display a RR that is also slightly lower than the control, they indicate large confidence limits. There is a noticeable difference in the risk reduction between Multifactorial Training, which places third in the treatment ranking, and Multifactorial Training (home), which has no clear positive effect on the reduction of fall risk. A similar difference but considerably smaller in magnitude is observed for Resistance Training and Resistance Training (home). These observations are in line with other meta-analyses. A possible explanation for these shared findings might be that the compliance, adherence, and the effort exerted in training are not sufficient when an intervention is performed largely or completely unsupervised at home.
The 2 interventions that fall out of line are the Postural Control Training and the Combined Postural Control Training. For these 2 approaches the unsupervised home training displays a larger effect than the supervised intervention or the same effect for postural control training and the combined intervention respectively. However, there are only 2 studies directly comparing Postural Control Training (home) with other interventions included in this analysis, and therefore, the findings should be interpreted with caution.
Conclusion
The evidence summarized in this network meta-analysis shows that balance training is the mode of physical activity or exercise that has the strongest positive impact on fall risk. For interventions that combine different modes of exercise or other non-exercise interventions the influence becomes less strong (up to about 12% reduction of fall risk) but still rather clear due to narrow confidence limits. With the exception of balance training–type exercise and the combined postural training approach, training performed at home was not as effective as training that was completely supervised. Future studies should investigate the role of training intensity and effort as well as the effects of multimodal exercise training over longer study periods up to ≥ 1 year.
This article has been excerpted from “Exercise Based Reduction of Falls in Community-dwelling Older Adults: A Network Meta-analysis” by the same authors, which was published in the European Review of Aging and Physical Activity. 2023;20(1):1. https://doi.org/10.1186/s11556-023-00311-w. Editing has occurred, including the renumbering of tables and figures, and references have been removed for brevity. Use is per CC Attribution 4.0 International License.
