After 65, a fall is the most common reason people lose their independence, and staying steady is trainable at no cost. The strongest lever you control is exercise that challenges your balance: pooled across 108 trials and 23,407 older adults, exercise cut the rate of falls 23%, balance and functional training 24%, and programs that both challenged balance and ran more than three hours a week cut falls 39%.
Ten minutes at a kitchen counter on most days reaches that dose with no equipment and no class.
The other half of the picture is not exercise at all:
- a sedative or an antidepressant
- blood pressure that drops when you stand
- a cataract
- a loose crystal in the inner ear a clinician can often fix in one appointment
- a dark route to the bathroom at night
Train the balance, then check the short list of fixable causes.
Findings & Outcomes
What It Is
Fall prevention trains the exact ability that fails when someone goes down: holding a position, shifting your weight onto one leg, and catching yourself before you hit the floor. It needs no equipment and no membership, and a kitchen counter to hold is the whole setup.
The cost of a fall after 65 is the reason to start early. The US Centers for Disease Control reports that about one in four adults aged 65 and over falls each year, and that over 95% of hip fractures come from a fall. In the three months after a hip fracture, the risk of dying runs about six times higher in women and eight times higher in men than in similar people without a fracture, and for many who survive it, independent living does not return.
Two levers do most of the work, and the evidence behind them differs in strength. The strongest thing you control is exercise that challenges your balance, backed by high-certainty trials, and adding leg strength does more. The second is a short list of causes outside the legs, several of them more fixable than balance is trainable.
What reduces falls
Two kinds of evidence sit in the ladder below, and they are not equally firm. Balance and functional training is the one exercise type with high-certainty trials behind it, and adding leg strength to it does more. Tai chi sits alongside it, not ahead of it. Below the exercise findings come the causes outside the legs: removing home hazards, reviewing sedating medication, checking blood pressure on standing, treating a cataract or a loose inner-ear crystal. The nulls belong in the ladder too, each at its own weight. Read it top to bottom; each card carries its finding at the strength of its own evidence.
The Research & Studies
Everything here is based on the research we have collected and checked, sorted into groups and ordered with the strongest evidence first. Click any claim to open the studies behind it.
Balance And Falls
Exercise cut the rate of falls 23% and the number who fell 15%
Pooling 108 trials and 23,407 older adults living at home, exercise of any kind cut how often people fell by 23% and cut the number who fell at all by 15%, both on high-certainty evidence. Fall-related fractures dropped about 27%, though that figure is less certain. These were supervised, supported programs, so a reader training alone at home is not getting quite the conditions the trials created.
Exercise reduced the rate of falls by 23% (rate ratio 0.77, 95% CI 0.71 to 0.83; 59 studies, 12,981 participants) and the number of people who fell at least once by 15% (risk ratio 0.85; 63 studies, 13,518 participants). Both high-certainty. Fall-related fractures fell about 27%, at low certainty. Measured in: 108 randomized trials across 25 countries, 23,407 community-dwelling participants, average age 76, 77% women. This review covers community-dwelling older adults and excludes residential care by design, so it says nothing either way about care homes. Adherence in the trials was supported and supervised in ways an unsupervised reader at home is not.
Who this may not transfer to:77% of participants were women, which is the reverse of the usual skew in exercise research and roughly matches who falls and who fractures. Men are the smaller group here and the estimate for them is less precise.
The study · 1
Sherrington et al., exercise for preventing falls in older people living in the community · Cochrane Database Syst Rev 2019;1(1):CD012424
Counts once: this finding and 4 others here come from the same source, so they are one body of evidence, not separate confirmations.
Balance and functional training cut the rate of falls 24%, the only high-certainty exercise type
Of all the kinds of exercise, the one with the firmest evidence is training that actually challenges your balance: standing in narrow or one-legged positions, moving without holding on, controlled turns. It cut the rate of falls by 24% across 39 trials, and it is the only exercise type here rated high-certainty. In the larger dose analysis, the programs that combined a real balance challenge with more than three hours a week did best.
Balance and functional exercise reduced the rate of falls by 24% (rate ratio 0.76, 95% CI 0.70 to 0.81; 39 studies, 7,920 participants), on high-certainty evidence. It is the only exercise type here with high-certainty evidence behind it. Measured in: 39 trials of balance and functional training in community-dwelling older adults, plus 88 trials and 19,478 participants in the 2017 dose analysis. The dose finding is a subgroup comparison across trials rather than a randomized comparison of doses, so the people who did three hours a week were also the people well enough to do three hours a week.
Who this may not transfer to:Sex is not broken out for this subgroup. The parent review was 77% women.
Aim your ten minutes at positions your balance has to hold, not at effort for its own sake, and build toward roughly three hours a week in total. The three-hours figure comes from comparing across trials rather than from a trial that assigned doses, so treat it as a target to grow into, not a threshold you have failed below.
The studies · 2
Sherrington et al., exercise for preventing falls in older people living in the community · Cochrane Database Syst Rev 2019;1(1):CD012424
Sherrington et al., exercise to prevent falls in older adults, updated systematic review and meta-analysis · Br J Sports Med 2017;51(24):1750-1758
Counts once: this finding and 4 others here come from the same source, so they are one body of evidence, not separate confirmations.
A nurse-led risk-assessment program did not cut serious fall injuries
When all of this was assembled into a single nurse-led program, risk assessment, an individual plan and referrals, for 5,451 adults over 70, the rate of a first serious fall injury did not drop significantly. Self-reported fall injuries were slightly lower.
The rate of a first adjudicated serious fall injury did not differ: 4.9 per 100 person-years in the intervention group against 5.3 in the control group (hazard ratio 0.92, 95% CI 0.80 to 1.06, P = 0.25). Participant-reported fall injuries were slightly lower (25.6 against 28.6 per 100 person-years, hazard ratio 0.90, 95% CI 0.83 to 0.99). Measured in: 5,451 community-dwelling adults aged 70 and over at increased risk of fall injury, in 86 primary care practices across 10 US health systems. Randomized by practice rather than by person, which reduces the effective sample below its headline size. The endpoint was time to first ADJUDICATED SERIOUS fall injury, which is a much higher bar than "did you fall", so this is not evidence that nothing changed at all. It is evidence that the serious-injury endpoint did not move.
Who this may not transfer to:About 62% women.
The result is about handing someone a plan versus them actually doing the balance training, getting the cataract done and the sedative reviewed. The individual levers still have their own evidence; what has not been shown is that you can bundle them and hand the bundle over.
The study · 1
Bhasin et al., a randomized trial of a multifactorial strategy to prevent serious fall injuries (STRIDE) · N Engl J Med 2020;383(2):129-140
Removing home hazards cut falls 26% overall and 38% in high-risk people, nothing in low-risk
Removing trip hazards in the home cut falls by 26% overall, and by 38% in people already at high risk of falling, but did nothing at all in unselected low-risk households. The versions that worked were delivered by an occupational therapist walking through the actual home, not by a checklist.
Home fall-hazard interventions reduced the overall rate of falls by 26% (rate ratio 0.74, 95% CI 0.61 to 0.91; 12 studies, 5,293 participants) at moderate certainty. In people already at high risk of falling the reduction was 38% (rate ratio 0.62, 95% CI 0.56 to 0.70; 9 studies) at high certainty. In unselected community groups there was no reduction at all (rate ratio 1.05, 95% CI 0.96 to 1.16). Measured in: 22 randomized trials in 10 countries, 8,463 community-dwelling older adults, average age 78, 65% women. Almost all of the benefit is concentrated in people who have already fallen or are otherwise at high risk, and the effective versions were delivered by an occupational therapist rather than by a checklist. Doing this to a low-risk household changed nothing.
Who this may not transfer to:65% women. Housing type and who lives alone differ by sex at this age, which plausibly changes both the hazards present and who acts on the advice.
This is worth doing if you have already fallen or are otherwise at high risk; in a low-risk household it changed nothing, so it is a targeted move rather than a universal one.
The study · 1
Clemson et al., environmental interventions for preventing falls in older people living in the community · Cochrane Database Syst Rev 2023;3(3):CD013258
Counts once: this finding and 1 other here come from the same source, so they are one body of evidence, not separate confirmations.
Balance plus resistance work cut falls 34%, the largest number in the review
Programs that added resistance work to balance training cut falls by 34%, the largest single number in the review. It also rests on the fewest trials, so the confidence interval is wide and the certainty only moderate.
Programs combining several exercise types reduced the rate of falls by 34% (rate ratio 0.66, 95% CI 0.50 to 0.88; 11 studies, 1,374 participants), on moderate-certainty evidence. Measured in: Community-dwelling older adults within the same Cochrane review of 108 trials. This is the largest point estimate in the review and it rests on the fewest trials, so the confidence interval is wide and the certainty is moderate rather than high. It should not be read as proof that adding weights beats balance work alone.
Who this may not transfer to:Sex is not broken out for this subgroup. The parent review was 77% women.
Keep balance work as the base and add leg strength for the other things it does, catching a stumble and building bone, rather than reading this as proof that weights beat balance training on their own.
The study · 1
Sherrington et al., exercise for preventing falls in older people living in the community · Cochrane Database Syst Rev 2019;1(1):CD012424
Counts once: this finding and 4 others here come from the same source, so they are one body of evidence, not separate confirmations.
Tai chi cut the rate of falls 19%, and made people 24% less likely to fall
Tai chi cut the rate of falls by 19% across seven trials on low-certainty evidence, and a later review of 24 trials found people were 24% less likely to become a faller, with more benefit at three or more sessions a week than at two. That puts it roughly level with other good balance training rather than ahead of it.
Tai Chi reduced the rate of falls by 19% (rate ratio 0.81, 95% CI 0.67 to 0.99; 7 studies, 2,655 participants) on low-certainty evidence. A later review of 24 randomized trials found a 24% lower risk of being a faller (risk ratio 0.76, 95% CI 0.71 to 0.82), with more effect at three or more sessions a week (risk ratio 0.67, 95% CI 0.58 to 0.79) than at two (0.78, 0.73 to 0.84). Measured in: 7 trials, 2,655 community-dwelling older adults in the Cochrane analysis; 24 trials in the 2023 review, predominantly older women. The Cochrane certainty grade for Tai Chi is low, and the pooled point estimate sits slightly behind general balance and functional training rather than ahead of it. Trials vary widely in style, form length and teacher, and the comparator is often stretching or usual care rather than another balance program.
Who this may not transfer to:Neither review aggregates a sex breakdown for the Tai Chi trials. Both describe samples weighted toward women.
The studies · 2
Sherrington et al., exercise for preventing falls in older people living in the community · Cochrane Database Syst Rev 2019;1(1):CD012424
Chen et al., Tai Chi for fall prevention and balance improvement in older adults · Front Public Health 2023;11:1236050
Counts once: this finding and 4 others here come from the same source, so they are one body of evidence, not separate confirmations.
Therapeutic tai ji quan cut falls 31% against a full multimodal exercise program
In the strongest single trial, 670 higher-risk adults over 70 trained for 24 weeks: a therapeutic tai ji quan program cut falls 58% against stretching and 31% against a full multimodal program of balance, aerobics, strength and flexibility. The multimodal program itself cut falls 40% against stretching.
Over 24 weeks, a therapeutic Tai Ji Quan program cut falls 58% against stretching (incidence rate ratio 0.42, 95% CI 0.31 to 0.56) and 31% against a multimodal exercise program of balance, aerobics, strength and flexibility (IRR 0.69, 95% CI 0.52 to 0.94). Multimodal exercise itself cut falls 40% against stretching (IRR 0.60, 95% CI 0.45 to 0.80). Measured in: 670 community-dwelling adults aged 70 and over in seven Oregon cities who had fallen in the past year or had impaired mobility. Mean age 77.7, 436 women (65%), 92% White. The trial that produced this is not inside the Cochrane pool, which searched only to May 2018 and lists it as awaiting classification, so those two really are separate. The other review cited alongside it does contain this trial and re-pools the Cochrane studies, so it is not a third independent confirmation. The trial author holds the license to the intervention being tested.
Who this may not transfer to:65% women and 92% White. The trial was not powered to compare the effect between men and women, and it says nothing about people outside that ethnic makeup.
This striking result sits above the pooled average across all tai chi trials, it was run in one US region with 92% White participants, and its lead author holds the license to the program tested, which he declares in the paper. Read it as a strong single trial, not as proof tai chi beats other balance training in general.
The study · 1
Li et al., therapeutic Tai Ji Quan vs multimodal exercise to prevent falls among older adults at high risk of falling · JAMA Intern Med 2018;178(10):1301-1310
Stopping fall-risk drugs on its own did not reduce falls
Trials that actually stopped fall-risk drugs, on their own, did not reduce falls, fallers or injuries across five trials and 1,305 people. Two independent reviews reach the same result.
Deprescribing as a standalone intervention did not change the rate of falls (rate ratio 0.98, 95% CI 0.63 to 1.51), the number of people who fell (risk ratio 1.04, 95% CI 0.86 to 1.26) or fall-related injuries (rate ratio 0.89, 95% CI 0.57 to 1.39) over 6 to 12 months. Measured in: Five randomized trials, 1,305 participants. Two independent reviews reach the same null, which is why this sits at moderate rather than preliminary. Withdrawal is hard to achieve and hard to sustain in these trials, so this measures the whole strategy of trying to stop the drugs rather than the effect of successfully being off them.
Who this may not transfer to:No sex breakdown is pooled. Prescribing patterns for the drug classes involved differ between men and women, so the mix of drugs withdrawn probably differed too.
A medication review is still worth asking for, especially if a sedative or antidepressant was started or changed recently, but it works as one part of a plan rather than as the whole of it, and these drugs are hard to stop and often restarted, which is part of why the trials came out flat.
The study · 1
Lee et al., deprescribing fall-risk increasing drugs (FRIDs) for the prevention of falls and fall-related complications · BMJ Open 2021;11(2):e035978
Blood pressure dropping on standing raised the odds of falling 73%
Blood pressure that drops when you stand, a fall of 20 points on the top number or 10 on the bottom within three minutes, was linked to a 73% higher odds of falling across 63 studies and about 50,000 people. It held whichever way the drop was defined or measured.
Orthostatic hypotension was associated with falls at an odds ratio of 1.73 (95% CI 1.50 to 1.99). The association held regardless of study population, design, quality, which definition of orthostatic hypotension was used, or how blood pressure was measured. Measured in: 63 studies and 51,800 individuals reviewed, 50 studies and 49,164 individuals pooled; 39 of the 63 were cross-sectional and 24 longitudinal. What could explain it instead: Orthostatic hypotension travels with the things that cause falls independently: dehydration, Parkinson's disease, diabetic autonomic neuropathy, frailty, and the same psychotropic and antihypertensive drugs listed above. It may be a marker of that whole cluster rather than the mechanism.. The pooled odds ratio is unadjusted, so some of it is the illness and medication that cause orthostatic hypotension rather than the blood-pressure drop itself.
Who this may not transfer to:No sex breakdown is pooled. Resting blood pressure and orthostatic response differ between men and women after 65, so the same threshold may not carry the same risk in both.
It is easy to check and rarely checked: lie down for five minutes, measure, then stand and measure again at one and three minutes. The pooled figure is unadjusted, so some of it is the illness and medication behind the drop rather than the drop itself, which is a reason to raise it with a doctor rather than self-treat.
The study · 1
Mol et al., orthostatic hypotension and falls in older adults, systematic review and meta-analysis · J Am Med Dir Assoc 2019;20(5):589-597
Higher-dose vitamin D produced more falls, not fewer
Vitamin D is the most recommended fall supplement, and the trials have moved against it. Taken daily, 2,000 and 4,000 IU a day produced more falls than 1,000, and a once-a-year 500,000 IU dose raised falls 15% and fractures 26%. No dose in any of these trials reduced falls.
Higher doses did worse. In a daily-dosing trial (STURDY) in older adults at elevated fall risk, 2,000 IU/day carried more falls than 1,000 IU/day (HR 1.54, 95% CI 1.01 to 2.34), and 4,000 IU/day trended the same way without reaching significance (HR 1.41, 95% CI 0.92 to 2.16). A separate trial giving 500,000 IU once a year for three to five years raised falls (IRR 1.15) and fractures (IRR 1.26). No regimen in any of these trials reduced falls. Measured in: Community-dwelling older adults across four trials and reviews, including a daily-dose trial that was 44% women and 56% men, an annual-megadose trial in women only, and a monthly-dose trial 67% women. The harm is not confined to exotic megadoses. The clearest dose-response evidence here comes from a trial giving vitamin D DAILY, where 2,000 and 4,000 IU a day did worse than 1,000. The annual 500,000 IU result is the most dramatic but it is also the least like anything a reader takes. The three-month timing detail inside that trial was a post hoc finding. None of this speaks to treating a diagnosed deficiency, which is a different question with a different answer.
Who this may not transfer to:Unusually for this literature, the daily-dose trial was male-majority at 56% men. The annual-megadose trial enrolled women only and the monthly-dose trial was 67% women, so the three do not describe the same population.
More is not better here, and the caution is not limited to dramatic annual megadoses; it shows up at the ordinary daily doses sold in bottles. Correcting a deficiency your doctor has measured is a separate question with a separate answer, and this says nothing about it.
The studies · 4
Sanders et al., annual high-dose oral vitamin D and falls and fractures in older women · JAMA 2010;303(18):1815-1822
Bischoff-Ferrari et al., monthly high-dose vitamin D treatment for the prevention of functional decline · JAMA Intern Med 2016;176(2):175-183
Appel et al., the effects of four doses of vitamin D supplements on falls in older adults (STURDY) · Ann Intern Med 2021;174(2):145-156
US Preventive Services Task Force, interventions to prevent falls in community-dwelling older adults · JAMA 2018;319(16):1696-1704
Fast-tracking first-eye cataract surgery cut the rate of falling 34%
Bringing first-eye cataract surgery forward to about four weeks instead of a twelve-month wait cut the rate of falling by 34% in 306 women over 70, with fewer fractures and better vision and general health.
Bringing first-eye cataract surgery forward to about four weeks rather than a twelve-month wait reduced the rate of falling by 34%, with fewer fractures and better visual function and general health status. Measured in: 306 women over 70 with cataract, randomized to expedited or routine surgery. The fall RATE fell while the proportion of people who fell at least once did not (49% of operated participants against 45% of controls), so this shifts repeat fallers rather than preventing first falls. The second-eye trial found a reduction of the same size that did not reach significance, so calling the second eye’s benefit smaller overstates what was shown.
Who this may not transfer to:No men were enrolled. Cataract surgery restores contrast sensitivity and depth judgement by the same optics in men, so the mechanism transfers, but the size of the fall reduction in older men has not been measured in a trial like this one.
The rate of falls dropped while the share of people who fell at least once did not, so this reduces repeat falls in people who keep falling rather than preventing a first fall. Treating a cataract that is affecting your sight is worth not putting off.
The study · 1
Harwood et al., falls and health status in elderly women following first eye cataract surgery · Br J Ophthalmol 2005;89(1):53-59
Single-lens distance glasses helped active wearers, but may have harmed the least active
Switching regular multifocal wearers to single-lens distance glasses for outdoor use cut falls in the people who were active outdoors, but among the least active it appeared to do harm, with outdoor falls rising. Multifocals blur the ground at exactly the distance where a curb sits.
Switching habitual multifocal wearers to single-lens glasses for outdoor use cut falls in the most active participants. Among the least active it may have done harm: outside falls rose (IRR 1.56, 95% CI 1.11 to 2.19), while all falls (1.29, 0.95 to 1.75) and injurious falls (1.34, 0.96 to 1.88) both crossed one. Measured in: 606 regular multifocal wearers, mean age 80, all at increased fall risk and using multifocals outdoors at least three times a week. Only the outside-falls endpoint reached significance in the least active group, and the trialists themselves put it as "may be harmful". The direction is consistent across all three endpoints, which makes it a signal rather than a settled finding.
Who this may not transfer to:65% women and 35% men.
If you are out and about a lot, a dedicated pair of distance glasses for walking is worth raising with your optometrist. If you are largely indoors and inactive, the swap may not help and could hurt, so this is not a blanket recommendation. The signal is consistent across the trial's endpoints but only the outdoor-falls one reached significance, so treat it as promising rather than settled.
The study · 1
Haran et al., effect on falls of providing single lens distance vision glasses to multifocal glasses wearers (VISIBLE) · BMJ 2010;340:c2265
Fear of falling carried 60% higher odds of a future fall
Being afraid of falling predicts falling: a single question about high versus low concern carried 60% higher odds of a future fall across 53 studies and 75,076 people, independent of previous falls. Oddly, the crude one-question measure predicted better than the detailed 16-item scales.
Concern about falling at baseline independently predicted falls at follow-up. A single-item measure of high versus low concern carried an odds ratio of 1.60 (95% CI 1.36 to 1.89). The 16-item Falls Efficacy Scale International carried 1.03 per point (1.02 to 1.05) and the 7-item short form 1.08 per point (1.05 to 1.11). Balance confidence measured by the ABC scale did not predict falls (0.97, 0.93 to 1.01). Measured in: 53 prospective studies, 75,076 participants, minimum six months of follow-up. Female representation ranged from 30% to 100%, median 60.3%. What could explain it instead: Concern about falling rises after a fall and rises with the impairments that cause falls, so some of this is the body reporting a real deficit rather than a fear causing one. The analyzes adjust for previous falls, which helps and does not settle it.. Twenty-six of the 53 studies were rated poor quality, though the associations held across quality strata. The measure that performs best is the crude single question, and the more elaborate scales perform worst, which is the opposite of what you would expect if the construct were cleanly defined.
Who this may not transfer to:Female representation across the included studies ranged from 30% to 100%, with a median of 60.3%, and some studies enrolled women only.
Some of that fear is a body accurately reporting a real deficit rather than an imagined one, which is a reason to answer it with training rather than with reassurance, since avoiding activity feeds the very loss of capacity that raises the risk.
The study · 1
Ellmers et al., does concern about falling predict future falls in older adults? A systematic review and meta-analysis · Age Ageing 2025;54(4):afaf089
Getting up at night to urinate raised falls about 1.2-fold
Getting up at night to urinate is linked to about a 1.2-fold rise in falls across nine studies, and a possible 1.3-fold rise in fractures. The reviewers rated it a moderate-quality predictor of falls and very low as a proven cause.
Nocturia was associated with about a 1.2-fold increase in falls (risk ratio 1.20, 95% CI 1.05 to 1.37, moderate quality) and a possible 1.3-fold increase in fractures (risk ratio 1.32, 95% CI 0.99 to 1.76, low quality). Measured in: Nine longitudinal observational studies of adults with and without nocturia. What could explain it instead: Nocturia is a symptom of heart failure, diabetes, prostate disease, sleep apnea and diuretic use, all of which raise fall risk on their own. It also tracks poor sleep, and sleeping badly impairs balance the next day.. The reviewers graded the evidence moderate for nocturia as a PREDICTOR of falls and very low for it as a CAUSE, which is the published answer to the obvious question. Gender was tested as an effect modifier and did not modify the association.
Who this may not transfer to:Gender was tested as an effect modifier and did not change the association.
The night trip stacks several risks at once: half asleep, low light, blood pressure at its lowest, often barefoot and in a hurry. A motion-sensor nightlight on the floor and sitting on the edge of the bed for fifteen seconds before standing remove most of it.
The study · 1
Pesonen et al., the impact of nocturia on falls and fractures, systematic review and meta-analysis · J Urol 2020;203(4):674-683
Multi-part programs cut the rate of falls 23 to 26%, though tailored ones left the number who fell unchanged
When several proven changes are combined into one program, falls drop. A fixed bundle of a few good components given to everyone cut the rate of falls by 26% and cut the number of people who fell by 18%, on moderate-certainty evidence. Programs that instead assess each person and hand back a tailored list cut the rate of falls by 23% but did not change how many people fell at all, on lower-certainty evidence.
Multifactorial programs, an individual risk assessment followed by a tailored set of actions, reduced the rate of falls by 23% (rate ratio 0.77, 95% CI 0.67 to 0.87; 19 studies, 5,853 participants) on low-certainty evidence, while the number of people who fell at least once was unchanged (risk ratio 0.96, 95% CI 0.90 to 1.03; 29 studies, 9,637 participants). Multiple-component programs, a fixed set of the same components given to everyone, reduced the rate of falls by 26% (rate ratio 0.74, 95% CI 0.60 to 0.91; 6 studies, 1,085 participants) and the number who fell by 18% (risk ratio 0.82, 95% CI 0.74 to 0.90; 11 studies, 1,980 participants), both on moderate-certainty evidence. Measured in: 62 randomized trials of community-dwelling older adults in the Cochrane review, spanning multifactorial and multiple-component designs. The two designs differ in a way that matters. Multifactorial means each person is assessed and then given their own mix of actions, and its fall-rate benefit is low-certainty while the number of people who fell did not move. Multiple-component means the same bundle of a few proven parts is delivered to everyone, and it carries moderate-certainty benefit on both measures. This sits alongside the STRIDE trial elsewhere on this page, a large newer multifactorial trial that assessed people and referred them onward and found no drop in serious fall injuries, which points to how much depends on the actions actually being carried out rather than recommended.
Who this may not transfer to:The review does not aggregate a single sex breakdown across its 62 trials. Fall-prevention trial populations in this literature run about two-thirds women, and the mix of risk factors addressed by a tailored program differs with the conditions more common in each sex.
The version with the firmer evidence is the small fixed set of components done by everyone, built around balance-challenging exercise. A personalized assessment is worth having, and its payoff depends on the recommended changes being carried out and kept up, which is where the tailored programs and the STRIDE trial on this page came out flat.
The study · 1
Hopewell et al., multifactorial and multiple component interventions for preventing falls in older people living in the community · Cochrane Database Syst Rev 2018;7(7):CD012221
Fast-stepping training cut the rate of falls 52%, on a thin base of 660 people
Training that provokes fast stepping, either stepping quickly to a cue or recovering from a deliberate shove or a treadmill slip, cut the rate of falls by 52% across seven trials, with the trials agreeing closely. It is a striking number on a thin base of 660 people.
Training that provokes fast stepping, either voluntary (stepping to a cue) or reactive (recovering from a shove or a treadmill slip), reduced the rate of falls by 52% (rate ratio 0.48, 95% CI 0.36 to 0.65) and the proportion of fallers by 49% (risk ratio 0.51, 95% CI 0.38 to 0.68), with no heterogeneity between trials. Measured in: Seven randomized trials, 660 older adults. 660 people across seven small trials is a thin base for an effect this large, and the reactive versions need a harness, a treadmill or a person to push you. What transfers to a kitchen counter is the voluntary stepping half.
Who this may not transfer to:The review does not aggregate sex. Stepping speed and reactive recovery differ with leg power, which differs by sex, so the size of the effect may not be the same in men and women.
The reactive versions need a harness, a treadmill or someone to push you, so the part that transfers to a kitchen counter is the voluntary half: stepping fast and accurately to a target on the floor.
The study · 1
Okubo, Schoene, Lord, step training improves reaction time, gait and balance and reduces falls in older people · Br J Sports Med 2017;51(7):586-593
Bundled vision programs made little or no difference to falls
Bundled vision-improvement programs made little or no difference to falls across three trials, on low-certainty evidence. That pools quite different things under one label, including new glasses prescriptions that carry their own adjustment risk, so it rules out very little and says nothing about treating a specific eye disease.
Vision improvement interventions made little or no difference to the rate of falls (rate ratio 1.12, 95% CI 0.84 to 1.50) or to the number of people who fell (risk ratio 1.09, 95% CI 0.79 to 1.50), on low-certainty evidence. Measured in: Three trials, 1,489 community-dwelling older adults within the Cochrane environmental review. Three trials and a wide interval, so this rules out very little. It pools different things under one label, including new spectacle prescriptions, which have their own adaptation risk, and it is not a verdict on treating a specific eye disease.
Who this may not transfer to:No separate sex breakdown for this subgroup. The parent review was 65% women.
The study · 1
Clemson et al., environmental interventions for preventing falls in older people living in the community · Cochrane Database Syst Rev 2023;3(3):CD013258
Counts once: this finding and 1 other here come from the same source, so they are one body of evidence, not separate confirmations.
A podiatry package cut the rate of falls 36% in people with painful feet
Painful feet and unsteady footwear are an overlooked part of falling. A twelve-month podiatry package, custom insoles, better-fitting shoes with a subsidy to buy them, a short daily foot-and-ankle exercise routine and fall-safety advice, cut the rate of falls by 36% in 305 older adults with disabling foot pain, and improved ankle strength, movement and standing balance along the way.
A twelve-month podiatry package cut the rate of falls by 36% (incidence rate ratio 0.64, 95% CI 0.45 to 0.91, P=0.01): 0.67 falls per person-year against 1.06 in the control group. The proportion who fell at least once did not differ (42% against 49%, P=0.19). At six months the package also improved ankle eversion strength, ankle range of motion and standing balance. Measured in: 305 community-dwelling older adults with disabling foot pain and an increased risk of falling, mean age 74, 69% women, in Melbourne, Australia. This is one trial in people who already had disabling foot pain, so it does not speak to feet that are comfortable. The package combined several parts, prefabricated orthoses, footwear advice with a purchase subsidy, a home foot-and-ankle exercise program three times a week, and a falls-education booklet, so it cannot separate which part carried the effect. As with several other levers here, the rate of falls dropped while the share of people who fell at least once did not, so it reduces repeat falls more than it prevents a first one.
Who this may not transfer to:69% women. Foot pain, footwear habits and the prevalence of conditions like bunions differ between older men and women, so the size of the benefit in a mostly-male group has not been measured in a trial like this one.
This was tested in people whose feet already hurt, so it is a targeted move for that situation rather than a universal one. It was a bundle, so no single piece can claim the credit, but the foot-and-ankle exercises and well-fitting flat, firm shoes are the parts easiest to act on at home. Painful feet are worth raising with a podiatrist for their own sake, and the steadier gait comes with it.
The study · 1
Spink et al., effectiveness of a multifaceted podiatry intervention to prevent falls in community dwelling older people with disabling foot pain · BMJ 2011;342:d3411
Resistance, dance and walking alone showed no fall reduction (too few trials to tell)
Pooled on their own, resistance training, walking and dance each showed no fall reduction. That is an absence of evidence rather than a sign they do not help: these were five small trials in 327 people for resistance, and the review says plainly they were too small to detect a difference and it never compared exercise types head to head. Balance-challenging exercise is simply the only type with high-certainty evidence for falls.
Pooled separately, none of these showed a fall reduction: resistance training rate ratio 1.14 (95% CI 0.67 to 1.97; 5 studies, 327 participants), dance 1.34 (0.98 to 1.83; 1 study), walking 1.14 (0.66 to 1.97; 2 studies). All are very-low-certainty evidence. Measured in: The smaller exercise-type subgroups within the same Cochrane review. This is an absence of evidence, not evidence of absence, and the difference matters. The review ran no meta-analysis comparing exercise types against each other and states that these trials were underpowered to detect differences between them. Five small trials in 327 people with I-squared of 67% cannot establish that strength work does not help. The defensible reading is that balance-challenging exercise is the only type with high-certainty evidence behind it for falls, and that resistance training alone has barely been tested for this particular outcome.
Who this may not transfer to:Sex is not broken out for these subgroups.
Do the strength work anyway, for leg power and bone, but make balance-challenging movement the part you count on for falls, because it is the part that has actually been shown to work.
The study · 1
Sherrington et al., exercise for preventing falls in older people living in the community · Cochrane Database Syst Rev 2019;1(1):CD012424
Counts once: this finding and 4 others here come from the same source, so they are one body of evidence, not separate confirmations.
Vestibular
One repositioning maneuver cleared positional vertigo in 56% against 21% sham
For the specific vertigo that makes the room spin for a few seconds when you roll over in bed or tip your head back, a repositioning maneuver clears it in about 56% of people against 21% with a sham. It treats one inner ear cause only.
The repositioning maneuver resolved vertigo in about 56% of people against 21% with a sham or no treatment. It treats posterior canal benign paroxysmal positional vertigo only. Measured in: Eleven mostly small randomized trials in adults with posterior canal benign paroxysmal positional vertigo. This does not treat dizziness in general. It does not treat the anterior or horizontal canal forms, and it does nothing for light-headedness on standing, inner-ear disease, or dizziness from medication or heart rhythm. If the room does not spin, and spin specifically when you move your head, this is not the problem it solves.
Who this may not transfer to:The review reports roughly 1 man to 1.5 women, across ages 18 to 90.
It does nothing for light-headedness on standing, general dizziness, or dizziness from medication or heart rhythm. The clue that it is the problem this solves is short bursts of true spinning triggered by head movement. Recurrence runs around a third, so it often returns, and the same maneuver treats it again.
The study · 1
Hilton and Pinder, the Epley (canalith repositioning) manoeuvre for benign paroxysmal positional vertigo · Cochrane Database Syst Rev 2014;(12):CD003162
Bone Density
After a hip fracture, death risk ran about 6 times higher in women and 8 in men
A hip fracture is why falls matter so much. In the three months after one, the risk of dying ran about 5.75 times higher in women and 7.95 times higher in men than in similar people without a fracture, and the excess narrows but persists for years.
In the first three months after a hip fracture, mortality ran 5.75 times higher in women and 7.95 times higher in men than in age- and sex-matched people without a fracture. The excess narrows afterwards but does not disappear. Measured in: Prospective cohort studies of older adults with hip fracture, published 1957 to May 2009. These are excess hazards relative to matched controls, not crude death rates, and the age-80 figures are reported for white women and men specifically. A hip fracture also selects for people who were already frailer, so this is not the fracture acting alone.
Who this may not transfer to:This is one of the few findings on this page reported separately for men and women, and the sexes differ: men carry the higher relative hazard and the higher excess mortality at every time point.
The study · 1
Haentjens et al., meta-analysis: excess mortality after hip fracture among older women and men · Ann Intern Med 2010;152(6):380-390
Heavy resistance and impact training raised spine bone density 2.9%
Eight months of twice-weekly, 30-minute supervised heavy lifting and impact work raised spine bone density 2.9% in postmenopausal women with thinning bones, against a 1.2% loss in the comparison group, and beat it on every strength and function measure, with one minor injury across the whole trial.
Eight months of twice-weekly, 30-minute supervised high-intensity resistance and impact training beat home-based low-intensity exercise for lumbar spine bone mineral density (+2.9% against -1.2%), femoral neck density (+0.3% against -1.9%), cortical thickness and every functional performance measure. Compliance was 92%, with one adverse event, a minor lower back spasm. Measured in: 101 postmenopausal women with osteopenia or osteoporosis, screened to exclude conditions and drugs affecting bone. Every session was supervised by staff trained to coach heavy lifting, and the authors are explicit that the safety finding applies under those conditions. Bone density is a surrogate: this trial was not sized to measure fractures.
Who this may not transfer to:No men were enrolled. The same group later ran a version in men with low bone mass, and the trial cited here says nothing about them. A man reading this has the mechanism and not the number.
Every session was supervised by staff trained to coach heavy lifting, and the authors tie the safety to that supervision. Bone density is a stand-in for fracture, which the trial was not sized to measure, so read this as strong evidence for the bone number under supervision.
The study · 1
Watson et al., high-intensity resistance and impact training improves bone mineral density and physical function in postmenopausal women (LIFTMOR) · J Bone Miner Res 2018;33(2):211-220
Progress Markers
Failing a 10-second one-leg stance nearly doubled the death rate over seven years
In 1,702 people aged 51 to 75, one in five could not hold a 10-second stand on one leg, and over about seven years 17.5% of them died against 4.6% of those who could, a gap that held after adjusting for age, sex, weight and illness.
20.4% of participants could not hold a 10-second one-legged stance. Over a median seven years, 17.5% of them died against 4.6% of those who could. Adjusted for age, sex, body mass index and comorbidities, the hazard ratio for all-cause mortality was 1.84 (95% CI 1.23 to 2.78). Measured in: 1,702 individuals aged 51 to 75, 68% men, assessed between 2008 and 2020 at a Brazilian exercise medicine clinic. What could explain it instead: People who cannot balance for ten seconds are more likely to have undiagnosed neurological disease, sarcopenia, obesity and inactivity, and the cohort comes from a self-selected clinic population rather than a community sample. Reverse causation is live: early illness degrades balance before it is diagnosed.. The outcome is death, not falls, and the authors say plainly that they could not control for recent falls or physical activity. A test failed today is a summary of everything that has already happened to the legs, the inner ear and the nervous system, so it forecasts well without implying that training the test changes the forecast.
Who this may not transfer to:68% men, which is the more usual skew in this literature. The absolute mortality rates will differ in a sample with more women, though the test itself is not sex-specific.
The outcome here is death, not falls, and the authors could not account for recent falls or activity. A test failed today summarizes what has already happened to the legs, inner ear and nervous system, so it forecasts well without meaning that drilling the test itself changes the forecast.
The study · 1
Araujo et al., successful 10-second one-legged stance performance predicts survival in middle-aged and older individuals · Br J Sports Med 2022;56(17):975-980
The timed up-and-go test missed most future fallers at the usual cutoff
The timed up-and-go test, standing, walking three meters, turning and sitting, is a popular fall screen, but at the usual 13.5-second cutoff it missed most people who went on to fall, catching only about 31%, and the review could not show the score predicts falls at all.
At a threshold of 13.5 seconds or more, the test had specificity 0.74 and sensitivity 0.31 across the 10 studies that could be pooled, out of 25 reviewed. The review’s own regression found the score was not a significant predictor of falls (OR 1.01, 95% CI 1.00 to 1.02, p=0.05). Measured in: 25 studies of community-dwelling older adults. Sensitivity of 0.31 means the test misses most people who go on to fall, so a fast time is close to meaningless as reassurance. A slow time is informative; a fast one is not. The pooled analysis rests on 10 studies, and the review could not show the score predicts falls at all.
Who this may not transfer to:No pooled sex breakdown. Walking speed norms differ by sex and height, so a single time threshold is unlikely to sit in the same place for everyone.
A slow time is worth taking seriously; a fast time is close to meaningless as reassurance, so do not read a good score as a clean bill of health.
The study · 1
Barry et al., is the timed up and go test a useful predictor of risk of falls in community-dwelling older adults? · BMC Geriatr 2014;14:14
How It Works
Anatomy of the Practice
1The first sessions
You put yourself in positions your balance has to actively hold, a narrow stance, weight on one leg, moving without holding on, and your body starts catching its own small corrections instead of a rail catching them. Nothing has changed structurally yet, but you learn where your limits are, which is information you did not have before.
2Over the first weeks
Trained on most days, balance, gait and the speed of a corrective step improve, and confidence on your feet grows with them. The trial programs ran two or three sessions a week for at least twelve weeks; over this span the balance itself begins to hold.
3Over months
Fall rates drop, and they keep dropping for as long as the training continues. Leg power built alongside the balance work turns a stumble into a recovery step, and loading heavy enough builds bone, so a fall is less likely to break something. The benefit tracks how regularly you train, and balance decays back within weeks of stopping.
Balance improves only when it is challenged.
Almost all of the measured benefit runs through that one mechanism. A program works when it puts you in positions your balance has to actively hold: a reduced base of support, weight shifted onto one leg, moving without holding on, controlled turns, stepping over things. Sitting on a machine and pressing a heavy weight does not qualify, however hard it is, which is why strength training on its own has not been shown to prevent falls even though it builds a stronger person. The dose that moved the trials was two or three sessions a week, and the subgroup with the largest effect combined a high challenge to balance with more than three hours of it a week.
For many readers, a page of exercises aims at the wrong thing. A large share of falls trace to something in a pill organizer, a blood pressure reading, an inner ear, or the eyes. Being afraid of falling is its own risk factor: it makes people cautious, they move less, and the capacity they were protecting drops.
How to Train
Ways to Do It
The trials delivered supervised group classes because that is how a trial runs. What the analysis identified as the active ingredient was the balance challenge and the total hours, and both survive being broken into short daily doses at home for nothing. Start at the first rung and add the others as they suit you. Set up so a wobble costs nothing: practice at a counter, in a doorway, or behind a heavy chair, so you can hold on the instant you need it.
Ten minutes on most days, no equipment. Work the positions that stress balance: feet together, then heel-to-toe, then a single-leg stance with one finger on the counter, then the same with your head turning side to side, which is what happens when someone calls your name in a shop. Progress by removing support: two hands, then one, then one finger, then nothing, dropping to a lower level only when the current one is easy. Ten minutes on most days beats an hour on Sunday, and this is the part that reaches the trial dose without a class.
Two sessions a week of sit-to-stands without using your arms, step-ups onto a stair, and carrying something heavy across the room, hard enough that the last repetitions are difficult. This is the strength half: strong legs catch a stumble before it becomes a fall, and heavy loading builds bone. Practice getting up off the floor too, once a month with someone in the house, because the time spent down after a fall predicts how badly it turns out.
A community or senior-center balance class is free in many places, and tai chi is a well-studied version of the same thing. Paying buys supervision: a teacher who corrects you, or, if you are unsteady enough that you would not attempt a single-leg stance beside a counter, a physiotherapist who assesses you first and progresses the work safely. That assessment is the paid end of this practice, and it is the right first step for anyone who has already fallen.
Go Deeper
- Resistance training: the strength and bone half, the loading that makes a fall less likely to break a bone.
- Tai chi and qi gong: the movement practice with the most randomized evidence for balance in older adults.
- Vitamin D: why the fall trials have moved against it, and what correcting a measured deficiency is and is not.
- Osteoporosis: the bone side of the same picture, how strong bone better withstands the impact of a fall.
- Yang Sheng: nourishing life, the preventive tradition this practice belongs to.
The Chinese Medicine View
Two organ systems carry unsteadiness in the tradition, and they describe different people. The Kidney governs the bones (腎主骨) and generates marrow, and the decline of Kidney essence with age is the classical account of legs that no longer hold. The Lingshu puts it directly: when the sea of marrow runs short, the head spins, the ears ring, the shins ache, the vision dims, and the person wants to lie down. As a description of what an unsteady 80-year-old reports, that is close.
The Liver governs the sinews (肝主筋), and Liver Blood nourishes them along with the eyes. Liver Blood deficiency gives cramping and trembling limbs, numbness, and blurred vision, unsteadiness read through the connective tissue rather than the skeleton, and it is the more common picture in someone younger or after blood loss.
On dizziness itself the tradition's physicians have disagreed for centuries. The Su Wen assigns all wind, shaking, and dizziness to the Liver; Zhu Danxi held that without Phlegm there is no dizziness; Zhang Jingyue held that without deficiency there is no dizziness. Each school points at a different cause of the same symptom, and each calls for a different treatment.
That frame describes the practice from the inside; the trials measured proprioception, gait, blood pressure, and vestibular function. The two accounts lie over the same body without translating into each other. Qi is not proprioception, and the sea of marrow is not the vestibular system.
The tradition also does not prescribe the same practice for everyone. The Su Wen's five taxations name prolonged standing as injuring the bones and prolonged walking as injuring the sinews, so for someone depleted, recovering from illness, or short of Qi and Blood, the guidance is short, frequent sessions and stopping before exhaustion. Someone with rising Liver Yang, whose dizziness comes with a red face, headache, and irritability, is settled and directed downward, so slow weighted stepping suits that pattern better than anything vigorous.
Tai chi belongs here as the tradition's own answer: a martial art from the Chen village lineage of around the seventeenth century, whose weight transfer and rooted stance contain the same elements the modern trials identify, which is why researchers built therapeutic versions of it. If you have a practitioner, this is a good thing to ask about, because the answer depends on your pattern.
Cautions For This Practice
Everything to be aware of is here, in one place. This practice suits most healthy people; a few situations call for real care.
Mind-acting drugs raise the odds of falling about 1.4 to 2-fold
Pooled odds of falling: antidepressants 1.57 (95% CI 1.43 to 1.74), antipsychotics 1.54 (1.28 to 1.85), benzodiazepines 1.42 (1.22 to 1.65), tricyclic antidepressants 1.41 (1.07 to 1.86), SSRIs 2.02 (1.85 to 2.20), long-acting benzodiazepines 1.81 (1.05 to 3.16), short-acting benzodiazepines 1.27 (1.04 to 1.56). An association, not a trial. The trials that actually removed the drugs found no fall reduction, so this cannot be read as "stopping them will help". Our note about SSRIs specifically is an inference from the drug classes involved rather than something the paper separates out.Seppala et al., fall-risk-increasing drugs, systematic review and meta-analysis II: psychotropics
Blood pressure drugs and serious falls: the evidence conflicts
The evidence conflicts. One US cohort (Tinetti 2014) found about 40% higher injurious-fall risk on antihypertensives at moderate intensity, while other analyzes have found neutral or even lower risk for several drug classes, so the evidence does not point one way. Two bodies of evidence point opposite ways, and neither is strong enough to settle the question. What both sides agree on is narrower and more actionable: a drop in blood pressure on standing raises fall risk, and that is measurable in a chair and a stand. Nobody should stop a blood pressure medicine on the strength of a web page.Tinetti et al., antihypertensive medications and serious fall injuries in a nationally representative sample of older adults
Train within reach of something solid at first
Practice at a kitchen counter, in a doorway, or behind a heavy chair, so you can hold on the instant you need it, and clear the floor of anything loose. Wear shoes with a back and a grip, or bare feet, rather than loose slippers or socks. Progress by taking a hand or a finger off, not by removing the support before the balance is there. If you are unsteady enough that you would not attempt a single-leg stance beside a counter, that is a referral to a physiotherapist, not a page to follow alone. Away from practice, the trip to the toilet at night is the highest-risk moment, half asleep in low light with blood pressure at its lowest, so a motion-sensor nightlight on the floor and sitting on the edge of the bed for a slow count of fifteen before standing remove most of that risk.
Get assessed after any fall, and talk to a doctor about certain ones
A previous fall is the strongest single predictor of the next one, and a fall is something to tell a doctor about. Talk to a doctor promptly before starting anything here if unsteadiness came on over days rather than years, if a fall involved loss of consciousness or you cannot account for it, which needs a cardiac assessment rather than a balance class, if there was a head strike while taking a blood thinner, or if there is new numbness in both feet. Unsteadiness that arrives suddenly with slurred speech, weakness or numbness down one side, double vision or a severe new headache is a stroke presentation and a reason to call emergency services.
Ask about medication, blood pressure on standing, vision and the inner ear
Much of what causes falls is not in the legs. A medication review is worth asking a doctor or pharmacist for, especially if a sedative, sleeping tablet, antidepressant or opioid was started or increased in the past few months, though it works as one part of a plan rather than on its own. Blood pressure that drops when you stand is easy to check and rarely checked: lie for five minutes, then measure standing at one and three minutes. Multifocal glasses blur the ground where a curb sits, so a dedicated pair of distance glasses for walking is worth raising with an optometrist, and an untreated cataract is worth treating, since bringing surgery forward has been shown to cut falls. Seconds of violent spinning when you roll over in bed or tip your head back points at the inner ear, which a clinician can often settle in one appointment. None of these is a reason to stop a prescribed medicine on the strength of a web page.
Heavy resistance and impact work, if bone is the concern
Loading heavy enough to build bone was safe and effective in the trial that showed it, and in that trial every session was supervised by staff trained to coach heavy lifting. If you have known osteoporosis, a previous vertebral fracture, or uncontrolled high blood pressure, arrange supervision before starting heavy impact work. Vitamin D is widely taken for falls and the trials have moved against it, with higher daily doses doing worse than lower ones; correcting a measured deficiency is a separate question for your doctor.
Start slow, be smart, read the research, and consult a professional if you have any concerns. This is here to inform your choice, not make it for you.
Common Questions
How do I know if I'm at risk of falling?
Three quick checks. Can you hold a heel-to-toe stance for 10 seconds? Can you stand from a chair without using your arms, and how many times in 30 seconds? Have you fallen in the past year? A previous fall is the strongest single predictor there is, and failing the heel-to-toe hold marks increased risk on the CDC screen. A 10-second single-leg stance shows a great deal about the legs and the nervous system in a short test, though it forecasts risk without fixing anything.
What is the best exercise to prevent falls?
Balance and functional training, which cut the rate of falls 24% across 39 trials at high certainty. Adding resistance work took it to 34%. The programs with the largest effect combined a high challenge to balance with more than three hours a week. Strength training on its own has not been shown to prevent falls, though it does other necessary work, including the leg power that catches a stumble and the loading that builds bone.
Does tai chi actually prevent falls?
Yes, by roughly the same amount as other good balance training. Pooled across seven trials it cut the rate of falls 19% on low-certainty evidence, against 24% for balance and functional training generally. One well-run head-to-head found a therapeutic tai ji quan program beat a multimodal exercise program by 31%, in a trial whose lead author holds the license to the intervention. Tai chi is a good answer here, and it is also the tradition's own.
If I do the exercises, do I still need a medication and blood pressure check?
Yes, and for a lot of readers that check matters more than the exercises. A large share of falls trace to a sedative, sleeping tablet or antidepressant, or to blood pressure that drops when you stand, rather than to weak legs. A medication review with a doctor or pharmacist and a lying-then-standing blood pressure reading are quick, and they address causes that no amount of balance work reaches.
Why do I feel dizzy when I stand up?
Often blood pressure lagging behind the change in posture, which carries an odds ratio of 1.73 for falls and is easy to measure: lie five minutes, then check standing at one and three minutes. If instead the room spins violently for a few seconds when you roll over in bed or tip your head back, that points to a loose crystal in the inner ear, and one repositioning maneuver in a clinic resolves it far more often than a sham does.
Does vitamin D prevent falls?
No, and the trials have moved against it. Taken daily, 2,000 and 4,000 IU a day produced more falls than 1,000, and a once-a-year high dose raised falls and fractures. No dose in these trials reduced falls. Correcting a measured deficiency is a separate question for your doctor, but taking vitamin D as a fall preventive is not supported.
Explore Related
Other pages this one connects to, by the evidence they share, the outcomes they touch, and the ground they cover.
How this connects
- Related
- How to measure it
All 31 sources on this page independently checked and cross-referenced.
Thomas Dehli, Founder & Editor, Sacred Lotus
Sacred Lotus has published Chinese medicine reference material since 2001. Integrative pages are held to the same standard as the herb and formula library: cite the source, grade the claim at its real strength, and say where the research has not looked. This page is educational and it is not medical advice. Last reviewed and updated August 9, 2026.
Evidence strength
How confidently the research supports a claim. Strength describes the evidence, not our endorsement.